libgomp.info revision 1.11 1 1.11 mrg This is libgomp.info, produced by makeinfo version 6.5 from
2 1.11 mrg libgomp.texi.
3 1.1 mrg
4 1.11 mrg Copyright (C) 2006-2018 Free Software Foundation, Inc.
5 1.1 mrg
6 1.1 mrg Permission is granted to copy, distribute and/or modify this document
7 1.3 mrg under the terms of the GNU Free Documentation License, Version 1.3 or
8 1.1 mrg any later version published by the Free Software Foundation; with the
9 1.1 mrg Invariant Sections being "Funding Free Software", the Front-Cover texts
10 1.1 mrg being (a) (see below), and with the Back-Cover Texts being (b) (see
11 1.1 mrg below). A copy of the license is included in the section entitled "GNU
12 1.1 mrg Free Documentation License".
13 1.1 mrg
14 1.1 mrg (a) The FSF's Front-Cover Text is:
15 1.1 mrg
16 1.1 mrg A GNU Manual
17 1.1 mrg
18 1.1 mrg (b) The FSF's Back-Cover Text is:
19 1.1 mrg
20 1.1 mrg You have freedom to copy and modify this GNU Manual, like GNU
21 1.11 mrg software. Copies published by the Free Software Foundation raise funds
22 1.11 mrg for GNU development.
23 1.1 mrg INFO-DIR-SECTION GNU Libraries
24 1.1 mrg START-INFO-DIR-ENTRY
25 1.5 mrg * libgomp: (libgomp). GNU Offloading and Multi Processing Runtime Library.
26 1.1 mrg END-INFO-DIR-ENTRY
27 1.1 mrg
28 1.5 mrg This manual documents libgomp, the GNU Offloading and Multi
29 1.5 mrg Processing Runtime library. This is the GNU implementation of the
30 1.5 mrg OpenMP and OpenACC APIs for parallel and accelerator programming in
31 1.5 mrg C/C++ and Fortran.
32 1.1 mrg
33 1.1 mrg Published by the Free Software Foundation 51 Franklin Street, Fifth
34 1.1 mrg Floor Boston, MA 02110-1301 USA
35 1.1 mrg
36 1.11 mrg Copyright (C) 2006-2018 Free Software Foundation, Inc.
37 1.1 mrg
38 1.1 mrg Permission is granted to copy, distribute and/or modify this document
39 1.3 mrg under the terms of the GNU Free Documentation License, Version 1.3 or
40 1.1 mrg any later version published by the Free Software Foundation; with the
41 1.1 mrg Invariant Sections being "Funding Free Software", the Front-Cover texts
42 1.1 mrg being (a) (see below), and with the Back-Cover Texts being (b) (see
43 1.1 mrg below). A copy of the license is included in the section entitled "GNU
44 1.1 mrg Free Documentation License".
45 1.1 mrg
46 1.1 mrg (a) The FSF's Front-Cover Text is:
47 1.1 mrg
48 1.1 mrg A GNU Manual
49 1.1 mrg
50 1.1 mrg (b) The FSF's Back-Cover Text is:
51 1.1 mrg
52 1.1 mrg You have freedom to copy and modify this GNU Manual, like GNU
53 1.11 mrg software. Copies published by the Free Software Foundation raise funds
54 1.11 mrg for GNU development.
55 1.1 mrg
56 1.1 mrg
57 1.1 mrg File: libgomp.info, Node: Top, Next: Enabling OpenMP, Up: (dir)
58 1.1 mrg
59 1.1 mrg Introduction
60 1.1 mrg ************
61 1.1 mrg
62 1.11 mrg This manual documents the usage of libgomp, the GNU Offloading and Multi
63 1.11 mrg Processing Runtime Library. This includes the GNU implementation of the
64 1.11 mrg OpenMP (http://www.openmp.org) Application Programming Interface (API)
65 1.11 mrg for multi-platform shared-memory parallel programming in C/C++ and
66 1.11 mrg Fortran, and the GNU implementation of the OpenACC
67 1.11 mrg (https://www.openacc.org) Application Programming Interface (API) for
68 1.5 mrg offloading of code to accelerator devices in C/C++ and Fortran.
69 1.5 mrg
70 1.5 mrg Originally, libgomp implemented the GNU OpenMP Runtime Library.
71 1.5 mrg Based on this, support for OpenACC and offloading (both OpenACC and
72 1.5 mrg OpenMP 4's target construct) has been added later on, and the library's
73 1.5 mrg name changed to GNU Offloading and Multi Processing Runtime Library.
74 1.1 mrg
75 1.1 mrg * Menu:
76 1.1 mrg
77 1.1 mrg * Enabling OpenMP:: How to enable OpenMP for your applications.
78 1.1 mrg * Runtime Library Routines:: The OpenMP runtime application programming
79 1.1 mrg interface.
80 1.1 mrg * Environment Variables:: Influencing runtime behavior with environment
81 1.1 mrg variables.
82 1.7 mrg * Enabling OpenACC:: How to enable OpenACC for your
83 1.7 mrg applications.
84 1.7 mrg * OpenACC Runtime Library Routines:: The OpenACC runtime application
85 1.7 mrg programming interface.
86 1.7 mrg * OpenACC Environment Variables:: Influencing OpenACC runtime behavior with
87 1.7 mrg environment variables.
88 1.7 mrg * CUDA Streams Usage:: Notes on the implementation of
89 1.7 mrg asynchronous operations.
90 1.7 mrg * OpenACC Library Interoperability:: OpenACC library interoperability with the
91 1.7 mrg NVIDIA CUBLAS library.
92 1.1 mrg * The libgomp ABI:: Notes on the external ABI presented by libgomp.
93 1.5 mrg * Reporting Bugs:: How to report bugs in the GNU Offloading and
94 1.5 mrg Multi Processing Runtime Library.
95 1.1 mrg * Copying:: GNU general public license says
96 1.1 mrg how you can copy and share libgomp.
97 1.1 mrg * GNU Free Documentation License::
98 1.1 mrg How you can copy and share this manual.
99 1.1 mrg * Funding:: How to help assure continued work for free
100 1.1 mrg software.
101 1.3 mrg * Library Index:: Index of this documentation.
102 1.1 mrg
103 1.1 mrg
104 1.1 mrg File: libgomp.info, Node: Enabling OpenMP, Next: Runtime Library Routines, Prev: Top, Up: Top
105 1.1 mrg
106 1.1 mrg 1 Enabling OpenMP
107 1.1 mrg *****************
108 1.1 mrg
109 1.1 mrg To activate the OpenMP extensions for C/C++ and Fortran, the
110 1.11 mrg compile-time flag '-fopenmp' must be specified. This enables the OpenMP
111 1.11 mrg directive '#pragma omp' in C/C++ and '!$omp' directives in free form,
112 1.11 mrg 'c$omp', '*$omp' and '!$omp' directives in fixed form, '!$' conditional
113 1.11 mrg compilation sentinels in free form and 'c$', '*$' and '!$' sentinels in
114 1.11 mrg fixed form, for Fortran. The flag also arranges for automatic linking
115 1.11 mrg of the OpenMP runtime library (*note Runtime Library Routines::).
116 1.11 mrg
117 1.11 mrg A complete description of all OpenMP directives accepted may be found
118 1.11 mrg in the OpenMP Application Program Interface (http://www.openmp.org)
119 1.11 mrg manual, version 4.5.
120 1.1 mrg
121 1.1 mrg
122 1.1 mrg File: libgomp.info, Node: Runtime Library Routines, Next: Environment Variables, Prev: Enabling OpenMP, Up: Top
123 1.1 mrg
124 1.1 mrg 2 Runtime Library Routines
125 1.1 mrg **************************
126 1.1 mrg
127 1.5 mrg The runtime routines described here are defined by Section 3 of the
128 1.7 mrg OpenMP specification in version 4.5. The routines are structured in
129 1.1 mrg following three parts:
130 1.1 mrg
131 1.5 mrg * Menu:
132 1.1 mrg
133 1.5 mrg Control threads, processors and the parallel environment. They have C
134 1.5 mrg linkage, and do not throw exceptions.
135 1.1 mrg
136 1.1 mrg * omp_get_active_level:: Number of active parallel regions
137 1.1 mrg * omp_get_ancestor_thread_num:: Ancestor thread ID
138 1.5 mrg * omp_get_cancellation:: Whether cancellation support is enabled
139 1.5 mrg * omp_get_default_device:: Get the default device for target regions
140 1.1 mrg * omp_get_dynamic:: Dynamic teams setting
141 1.1 mrg * omp_get_level:: Number of parallel regions
142 1.3 mrg * omp_get_max_active_levels:: Maximum number of active regions
143 1.7 mrg * omp_get_max_task_priority:: Maximum task priority value that can be set
144 1.3 mrg * omp_get_max_threads:: Maximum number of threads of parallel region
145 1.1 mrg * omp_get_nested:: Nested parallel regions
146 1.5 mrg * omp_get_num_devices:: Number of target devices
147 1.1 mrg * omp_get_num_procs:: Number of processors online
148 1.5 mrg * omp_get_num_teams:: Number of teams
149 1.1 mrg * omp_get_num_threads:: Size of the active team
150 1.5 mrg * omp_get_proc_bind:: Whether theads may be moved between CPUs
151 1.1 mrg * omp_get_schedule:: Obtain the runtime scheduling method
152 1.5 mrg * omp_get_team_num:: Get team number
153 1.1 mrg * omp_get_team_size:: Number of threads in a team
154 1.3 mrg * omp_get_thread_limit:: Maximum number of threads
155 1.1 mrg * omp_get_thread_num:: Current thread ID
156 1.1 mrg * omp_in_parallel:: Whether a parallel region is active
157 1.3 mrg * omp_in_final:: Whether in final or included task region
158 1.5 mrg * omp_is_initial_device:: Whether executing on the host device
159 1.5 mrg * omp_set_default_device:: Set the default device for target regions
160 1.1 mrg * omp_set_dynamic:: Enable/disable dynamic teams
161 1.1 mrg * omp_set_max_active_levels:: Limits the number of active parallel regions
162 1.1 mrg * omp_set_nested:: Enable/disable nested parallel regions
163 1.1 mrg * omp_set_num_threads:: Set upper team size limit
164 1.1 mrg * omp_set_schedule:: Set the runtime scheduling method
165 1.1 mrg
166 1.5 mrg Initialize, set, test, unset and destroy simple and nested locks.
167 1.1 mrg
168 1.1 mrg * omp_init_lock:: Initialize simple lock
169 1.1 mrg * omp_set_lock:: Wait for and set simple lock
170 1.1 mrg * omp_test_lock:: Test and set simple lock if available
171 1.1 mrg * omp_unset_lock:: Unset simple lock
172 1.1 mrg * omp_destroy_lock:: Destroy simple lock
173 1.1 mrg * omp_init_nest_lock:: Initialize nested lock
174 1.1 mrg * omp_set_nest_lock:: Wait for and set simple lock
175 1.1 mrg * omp_test_nest_lock:: Test and set nested lock if available
176 1.1 mrg * omp_unset_nest_lock:: Unset nested lock
177 1.1 mrg * omp_destroy_nest_lock:: Destroy nested lock
178 1.1 mrg
179 1.5 mrg Portable, thread-based, wall clock timer.
180 1.1 mrg
181 1.1 mrg * omp_get_wtick:: Get timer precision.
182 1.1 mrg * omp_get_wtime:: Elapsed wall clock time.
183 1.1 mrg
184 1.1 mrg
185 1.1 mrg File: libgomp.info, Node: omp_get_active_level, Next: omp_get_ancestor_thread_num, Up: Runtime Library Routines
186 1.1 mrg
187 1.11 mrg 2.1 'omp_get_active_level' - Number of parallel regions
188 1.1 mrg =======================================================
189 1.1 mrg
190 1.1 mrg _Description_:
191 1.1 mrg This function returns the nesting level for the active parallel
192 1.1 mrg blocks, which enclose the calling call.
193 1.1 mrg
194 1.1 mrg _C/C++_
195 1.11 mrg _Prototype_: 'int omp_get_active_level(void);'
196 1.1 mrg
197 1.1 mrg _Fortran_:
198 1.11 mrg _Interface_: 'integer function omp_get_active_level()'
199 1.1 mrg
200 1.1 mrg _See also_:
201 1.1 mrg *note omp_get_level::, *note omp_get_max_active_levels::, *note
202 1.1 mrg omp_set_max_active_levels::
203 1.1 mrg
204 1.1 mrg _Reference_:
205 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.20.
206 1.1 mrg
207 1.1 mrg
208 1.5 mrg File: libgomp.info, Node: omp_get_ancestor_thread_num, Next: omp_get_cancellation, Prev: omp_get_active_level, Up: Runtime Library Routines
209 1.1 mrg
210 1.11 mrg 2.2 'omp_get_ancestor_thread_num' - Ancestor thread ID
211 1.1 mrg ======================================================
212 1.1 mrg
213 1.1 mrg _Description_:
214 1.1 mrg This function returns the thread identification number for the
215 1.5 mrg given nesting level of the current thread. For values of LEVEL
216 1.11 mrg outside zero to 'omp_get_level' -1 is returned; if LEVEL is
217 1.11 mrg 'omp_get_level' the result is identical to 'omp_get_thread_num'.
218 1.1 mrg
219 1.1 mrg _C/C++_
220 1.11 mrg _Prototype_: 'int omp_get_ancestor_thread_num(int level);'
221 1.1 mrg
222 1.1 mrg _Fortran_:
223 1.11 mrg _Interface_: 'integer function omp_get_ancestor_thread_num(level)'
224 1.11 mrg 'integer level'
225 1.1 mrg
226 1.1 mrg _See also_:
227 1.1 mrg *note omp_get_level::, *note omp_get_thread_num::, *note
228 1.1 mrg omp_get_team_size::
229 1.1 mrg
230 1.1 mrg _Reference_:
231 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.18.
232 1.1 mrg
233 1.1 mrg
234 1.5 mrg File: libgomp.info, Node: omp_get_cancellation, Next: omp_get_default_device, Prev: omp_get_ancestor_thread_num, Up: Runtime Library Routines
235 1.5 mrg
236 1.11 mrg 2.3 'omp_get_cancellation' - Whether cancellation support is enabled
237 1.5 mrg ====================================================================
238 1.5 mrg
239 1.5 mrg _Description_:
240 1.11 mrg This function returns 'true' if cancellation is activated, 'false'
241 1.11 mrg otherwise. Here, 'true' and 'false' represent their
242 1.11 mrg language-specific counterparts. Unless 'OMP_CANCELLATION' is set
243 1.5 mrg true, cancellations are deactivated.
244 1.5 mrg
245 1.5 mrg _C/C++_:
246 1.11 mrg _Prototype_: 'int omp_get_cancellation(void);'
247 1.5 mrg
248 1.5 mrg _Fortran_:
249 1.11 mrg _Interface_: 'logical function omp_get_cancellation()'
250 1.5 mrg
251 1.5 mrg _See also_:
252 1.5 mrg *note OMP_CANCELLATION::
253 1.5 mrg
254 1.5 mrg _Reference_:
255 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.9.
256 1.5 mrg
257 1.5 mrg
258 1.5 mrg File: libgomp.info, Node: omp_get_default_device, Next: omp_get_dynamic, Prev: omp_get_cancellation, Up: Runtime Library Routines
259 1.5 mrg
260 1.11 mrg 2.4 'omp_get_default_device' - Get the default device for target regions
261 1.5 mrg ========================================================================
262 1.5 mrg
263 1.5 mrg _Description_:
264 1.5 mrg Get the default device for target regions without device clause.
265 1.5 mrg
266 1.5 mrg _C/C++_:
267 1.11 mrg _Prototype_: 'int omp_get_default_device(void);'
268 1.1 mrg
269 1.5 mrg _Fortran_:
270 1.11 mrg _Interface_: 'integer function omp_get_default_device()'
271 1.5 mrg
272 1.5 mrg _See also_:
273 1.5 mrg *note OMP_DEFAULT_DEVICE::, *note omp_set_default_device::
274 1.5 mrg
275 1.5 mrg _Reference_:
276 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.30.
277 1.5 mrg
278 1.5 mrg
279 1.5 mrg File: libgomp.info, Node: omp_get_dynamic, Next: omp_get_level, Prev: omp_get_default_device, Up: Runtime Library Routines
280 1.5 mrg
281 1.11 mrg 2.5 'omp_get_dynamic' - Dynamic teams setting
282 1.1 mrg =============================================
283 1.1 mrg
284 1.1 mrg _Description_:
285 1.11 mrg This function returns 'true' if enabled, 'false' otherwise. Here,
286 1.11 mrg 'true' and 'false' represent their language-specific counterparts.
287 1.1 mrg
288 1.1 mrg The dynamic team setting may be initialized at startup by the
289 1.11 mrg 'OMP_DYNAMIC' environment variable or at runtime using
290 1.11 mrg 'omp_set_dynamic'. If undefined, dynamic adjustment is disabled by
291 1.11 mrg default.
292 1.1 mrg
293 1.1 mrg _C/C++_:
294 1.11 mrg _Prototype_: 'int omp_get_dynamic(void);'
295 1.1 mrg
296 1.1 mrg _Fortran_:
297 1.11 mrg _Interface_: 'logical function omp_get_dynamic()'
298 1.1 mrg
299 1.1 mrg _See also_:
300 1.1 mrg *note omp_set_dynamic::, *note OMP_DYNAMIC::
301 1.1 mrg
302 1.1 mrg _Reference_:
303 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.8.
304 1.1 mrg
305 1.1 mrg
306 1.1 mrg File: libgomp.info, Node: omp_get_level, Next: omp_get_max_active_levels, Prev: omp_get_dynamic, Up: Runtime Library Routines
307 1.1 mrg
308 1.11 mrg 2.6 'omp_get_level' - Obtain the current nesting level
309 1.1 mrg ======================================================
310 1.1 mrg
311 1.1 mrg _Description_:
312 1.1 mrg This function returns the nesting level for the parallel blocks,
313 1.1 mrg which enclose the calling call.
314 1.1 mrg
315 1.1 mrg _C/C++_
316 1.11 mrg _Prototype_: 'int omp_get_level(void);'
317 1.1 mrg
318 1.1 mrg _Fortran_:
319 1.11 mrg _Interface_: 'integer function omp_level()'
320 1.1 mrg
321 1.1 mrg _See also_:
322 1.1 mrg *note omp_get_active_level::
323 1.1 mrg
324 1.1 mrg _Reference_:
325 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.17.
326 1.1 mrg
327 1.1 mrg
328 1.7 mrg File: libgomp.info, Node: omp_get_max_active_levels, Next: omp_get_max_task_priority, Prev: omp_get_level, Up: Runtime Library Routines
329 1.1 mrg
330 1.11 mrg 2.7 'omp_get_max_active_levels' - Maximum number of active regions
331 1.1 mrg ==================================================================
332 1.1 mrg
333 1.1 mrg _Description_:
334 1.3 mrg This function obtains the maximum allowed number of nested, active
335 1.3 mrg parallel regions.
336 1.1 mrg
337 1.1 mrg _C/C++_
338 1.11 mrg _Prototype_: 'int omp_get_max_active_levels(void);'
339 1.1 mrg
340 1.1 mrg _Fortran_:
341 1.11 mrg _Interface_: 'integer function omp_get_max_active_levels()'
342 1.1 mrg
343 1.1 mrg _See also_:
344 1.1 mrg *note omp_set_max_active_levels::, *note omp_get_active_level::
345 1.1 mrg
346 1.1 mrg _Reference_:
347 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.16.
348 1.1 mrg
349 1.1 mrg
350 1.7 mrg File: libgomp.info, Node: omp_get_max_task_priority, Next: omp_get_max_threads, Prev: omp_get_max_active_levels, Up: Runtime Library Routines
351 1.7 mrg
352 1.11 mrg 2.8 'omp_get_max_task_priority' - Maximum priority value
353 1.7 mrg ========================================================
354 1.7 mrg
355 1.7 mrg that can be set for tasks.
356 1.7 mrg _Description_:
357 1.7 mrg This function obtains the maximum allowed priority number for
358 1.7 mrg tasks.
359 1.7 mrg
360 1.7 mrg _C/C++_
361 1.11 mrg _Prototype_: 'int omp_get_max_task_priority(void);'
362 1.1 mrg
363 1.7 mrg _Fortran_:
364 1.11 mrg _Interface_: 'integer function omp_get_max_task_priority()'
365 1.7 mrg
366 1.7 mrg _Reference_:
367 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.29.
368 1.7 mrg
369 1.7 mrg
370 1.7 mrg File: libgomp.info, Node: omp_get_max_threads, Next: omp_get_nested, Prev: omp_get_max_task_priority, Up: Runtime Library Routines
371 1.7 mrg
372 1.11 mrg 2.9 'omp_get_max_threads' - Maximum number of threads of parallel region
373 1.1 mrg ========================================================================
374 1.1 mrg
375 1.1 mrg _Description_:
376 1.3 mrg Return the maximum number of threads used for the current parallel
377 1.11 mrg region that does not use the clause 'num_threads'.
378 1.1 mrg
379 1.1 mrg _C/C++_:
380 1.11 mrg _Prototype_: 'int omp_get_max_threads(void);'
381 1.1 mrg
382 1.1 mrg _Fortran_:
383 1.11 mrg _Interface_: 'integer function omp_get_max_threads()'
384 1.1 mrg
385 1.1 mrg _See also_:
386 1.1 mrg *note omp_set_num_threads::, *note omp_set_dynamic::, *note
387 1.1 mrg omp_get_thread_limit::
388 1.1 mrg
389 1.1 mrg _Reference_:
390 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.3.
391 1.1 mrg
392 1.1 mrg
393 1.5 mrg File: libgomp.info, Node: omp_get_nested, Next: omp_get_num_devices, Prev: omp_get_max_threads, Up: Runtime Library Routines
394 1.1 mrg
395 1.11 mrg 2.10 'omp_get_nested' - Nested parallel regions
396 1.7 mrg ===============================================
397 1.1 mrg
398 1.1 mrg _Description_:
399 1.11 mrg This function returns 'true' if nested parallel regions are
400 1.11 mrg enabled, 'false' otherwise. Here, 'true' and 'false' represent
401 1.1 mrg their language-specific counterparts.
402 1.1 mrg
403 1.1 mrg Nested parallel regions may be initialized at startup by the
404 1.11 mrg 'OMP_NESTED' environment variable or at runtime using
405 1.11 mrg 'omp_set_nested'. If undefined, nested parallel regions are
406 1.1 mrg disabled by default.
407 1.1 mrg
408 1.1 mrg _C/C++_:
409 1.11 mrg _Prototype_: 'int omp_get_nested(void);'
410 1.1 mrg
411 1.1 mrg _Fortran_:
412 1.11 mrg _Interface_: 'logical function omp_get_nested()'
413 1.1 mrg
414 1.1 mrg _See also_:
415 1.1 mrg *note omp_set_nested::, *note OMP_NESTED::
416 1.1 mrg
417 1.1 mrg _Reference_:
418 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.11.
419 1.1 mrg
420 1.1 mrg
421 1.5 mrg File: libgomp.info, Node: omp_get_num_devices, Next: omp_get_num_procs, Prev: omp_get_nested, Up: Runtime Library Routines
422 1.1 mrg
423 1.11 mrg 2.11 'omp_get_num_devices' - Number of target devices
424 1.1 mrg =====================================================
425 1.1 mrg
426 1.1 mrg _Description_:
427 1.5 mrg Returns the number of target devices.
428 1.5 mrg
429 1.5 mrg _C/C++_:
430 1.11 mrg _Prototype_: 'int omp_get_num_devices(void);'
431 1.5 mrg
432 1.5 mrg _Fortran_:
433 1.11 mrg _Interface_: 'integer function omp_get_num_devices()'
434 1.5 mrg
435 1.5 mrg _Reference_:
436 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.31.
437 1.5 mrg
438 1.5 mrg
439 1.5 mrg File: libgomp.info, Node: omp_get_num_procs, Next: omp_get_num_teams, Prev: omp_get_num_devices, Up: Runtime Library Routines
440 1.5 mrg
441 1.11 mrg 2.12 'omp_get_num_procs' - Number of processors online
442 1.5 mrg ======================================================
443 1.5 mrg
444 1.5 mrg _Description_:
445 1.5 mrg Returns the number of processors online on that device.
446 1.1 mrg
447 1.1 mrg _C/C++_:
448 1.11 mrg _Prototype_: 'int omp_get_num_procs(void);'
449 1.1 mrg
450 1.1 mrg _Fortran_:
451 1.11 mrg _Interface_: 'integer function omp_get_num_procs()'
452 1.1 mrg
453 1.1 mrg _Reference_:
454 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.5.
455 1.1 mrg
456 1.1 mrg
457 1.5 mrg File: libgomp.info, Node: omp_get_num_teams, Next: omp_get_num_threads, Prev: omp_get_num_procs, Up: Runtime Library Routines
458 1.1 mrg
459 1.11 mrg 2.13 'omp_get_num_teams' - Number of teams
460 1.5 mrg ==========================================
461 1.1 mrg
462 1.1 mrg _Description_:
463 1.5 mrg Returns the number of teams in the current team region.
464 1.5 mrg
465 1.5 mrg _C/C++_:
466 1.11 mrg _Prototype_: 'int omp_get_num_teams(void);'
467 1.5 mrg
468 1.5 mrg _Fortran_:
469 1.11 mrg _Interface_: 'integer function omp_get_num_teams()'
470 1.5 mrg
471 1.5 mrg _Reference_:
472 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.32.
473 1.5 mrg
474 1.5 mrg
475 1.5 mrg File: libgomp.info, Node: omp_get_num_threads, Next: omp_get_proc_bind, Prev: omp_get_num_teams, Up: Runtime Library Routines
476 1.5 mrg
477 1.11 mrg 2.14 'omp_get_num_threads' - Size of the active team
478 1.5 mrg ====================================================
479 1.5 mrg
480 1.5 mrg _Description_:
481 1.11 mrg Returns the number of threads in the current team. In a sequential
482 1.11 mrg section of the program 'omp_get_num_threads' returns 1.
483 1.1 mrg
484 1.1 mrg The default team size may be initialized at startup by the
485 1.11 mrg 'OMP_NUM_THREADS' environment variable. At runtime, the size of
486 1.11 mrg the current team may be set either by the 'NUM_THREADS' clause or
487 1.11 mrg by 'omp_set_num_threads'. If none of the above were used to define
488 1.11 mrg a specific value and 'OMP_DYNAMIC' is disabled, one thread per CPU
489 1.11 mrg online is used.
490 1.1 mrg
491 1.1 mrg _C/C++_:
492 1.11 mrg _Prototype_: 'int omp_get_num_threads(void);'
493 1.1 mrg
494 1.1 mrg _Fortran_:
495 1.11 mrg _Interface_: 'integer function omp_get_num_threads()'
496 1.1 mrg
497 1.1 mrg _See also_:
498 1.1 mrg *note omp_get_max_threads::, *note omp_set_num_threads::, *note
499 1.1 mrg OMP_NUM_THREADS::
500 1.1 mrg
501 1.1 mrg _Reference_:
502 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.2.
503 1.5 mrg
504 1.5 mrg
505 1.5 mrg File: libgomp.info, Node: omp_get_proc_bind, Next: omp_get_schedule, Prev: omp_get_num_threads, Up: Runtime Library Routines
506 1.5 mrg
507 1.11 mrg 2.15 'omp_get_proc_bind' - Whether theads may be moved between CPUs
508 1.5 mrg ===================================================================
509 1.5 mrg
510 1.5 mrg _Description_:
511 1.11 mrg This functions returns the currently active thread affinity policy,
512 1.11 mrg which is set via 'OMP_PROC_BIND'. Possible values are
513 1.11 mrg 'omp_proc_bind_false', 'omp_proc_bind_true',
514 1.11 mrg 'omp_proc_bind_master', 'omp_proc_bind_close' and
515 1.11 mrg 'omp_proc_bind_spread'.
516 1.5 mrg
517 1.5 mrg _C/C++_:
518 1.11 mrg _Prototype_: 'omp_proc_bind_t omp_get_proc_bind(void);'
519 1.5 mrg
520 1.5 mrg _Fortran_:
521 1.11 mrg _Interface_: 'integer(kind=omp_proc_bind_kind) function
522 1.11 mrg omp_get_proc_bind()'
523 1.5 mrg
524 1.5 mrg _See also_:
525 1.5 mrg *note OMP_PROC_BIND::, *note OMP_PLACES::, *note
526 1.5 mrg GOMP_CPU_AFFINITY::,
527 1.5 mrg
528 1.5 mrg _Reference_:
529 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.22.
530 1.1 mrg
531 1.1 mrg
532 1.5 mrg File: libgomp.info, Node: omp_get_schedule, Next: omp_get_team_num, Prev: omp_get_proc_bind, Up: Runtime Library Routines
533 1.1 mrg
534 1.11 mrg 2.16 'omp_get_schedule' - Obtain the runtime scheduling method
535 1.1 mrg ==============================================================
536 1.1 mrg
537 1.1 mrg _Description_:
538 1.5 mrg Obtain the runtime scheduling method. The KIND argument will be
539 1.11 mrg set to the value 'omp_sched_static', 'omp_sched_dynamic',
540 1.11 mrg 'omp_sched_guided' or 'omp_sched_auto'. The second argument,
541 1.7 mrg CHUNK_SIZE, is set to the chunk size.
542 1.1 mrg
543 1.1 mrg _C/C++_
544 1.11 mrg _Prototype_: 'void omp_get_schedule(omp_sched_t *kind, int
545 1.11 mrg *chunk_size);'
546 1.1 mrg
547 1.1 mrg _Fortran_:
548 1.11 mrg _Interface_: 'subroutine omp_get_schedule(kind, chunk_size)'
549 1.11 mrg 'integer(kind=omp_sched_kind) kind'
550 1.11 mrg 'integer chunk_size'
551 1.1 mrg
552 1.1 mrg _See also_:
553 1.1 mrg *note omp_set_schedule::, *note OMP_SCHEDULE::
554 1.1 mrg
555 1.1 mrg _Reference_:
556 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.13.
557 1.5 mrg
558 1.5 mrg
559 1.5 mrg File: libgomp.info, Node: omp_get_team_num, Next: omp_get_team_size, Prev: omp_get_schedule, Up: Runtime Library Routines
560 1.5 mrg
561 1.11 mrg 2.17 'omp_get_team_num' - Get team number
562 1.5 mrg =========================================
563 1.5 mrg
564 1.5 mrg _Description_:
565 1.5 mrg Returns the team number of the calling thread.
566 1.5 mrg
567 1.5 mrg _C/C++_:
568 1.11 mrg _Prototype_: 'int omp_get_team_num(void);'
569 1.5 mrg
570 1.5 mrg _Fortran_:
571 1.11 mrg _Interface_: 'integer function omp_get_team_num()'
572 1.5 mrg
573 1.5 mrg _Reference_:
574 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.33.
575 1.1 mrg
576 1.1 mrg
577 1.5 mrg File: libgomp.info, Node: omp_get_team_size, Next: omp_get_thread_limit, Prev: omp_get_team_num, Up: Runtime Library Routines
578 1.1 mrg
579 1.11 mrg 2.18 'omp_get_team_size' - Number of threads in a team
580 1.1 mrg ======================================================
581 1.1 mrg
582 1.1 mrg _Description_:
583 1.1 mrg This function returns the number of threads in a thread team to
584 1.5 mrg which either the current thread or its ancestor belongs. For
585 1.11 mrg values of LEVEL outside zero to 'omp_get_level', -1 is returned; if
586 1.11 mrg LEVEL is zero, 1 is returned, and for 'omp_get_level', the result
587 1.11 mrg is identical to 'omp_get_num_threads'.
588 1.1 mrg
589 1.1 mrg _C/C++_:
590 1.11 mrg _Prototype_: 'int omp_get_team_size(int level);'
591 1.1 mrg
592 1.1 mrg _Fortran_:
593 1.11 mrg _Interface_: 'integer function omp_get_team_size(level)'
594 1.11 mrg 'integer level'
595 1.1 mrg
596 1.1 mrg _See also_:
597 1.1 mrg *note omp_get_num_threads::, *note omp_get_level::, *note
598 1.1 mrg omp_get_ancestor_thread_num::
599 1.1 mrg
600 1.1 mrg _Reference_:
601 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.19.
602 1.1 mrg
603 1.1 mrg
604 1.1 mrg File: libgomp.info, Node: omp_get_thread_limit, Next: omp_get_thread_num, Prev: omp_get_team_size, Up: Runtime Library Routines
605 1.1 mrg
606 1.11 mrg 2.19 'omp_get_thread_limit' - Maximum number of threads
607 1.1 mrg =======================================================
608 1.1 mrg
609 1.1 mrg _Description_:
610 1.3 mrg Return the maximum number of threads of the program.
611 1.1 mrg
612 1.1 mrg _C/C++_:
613 1.11 mrg _Prototype_: 'int omp_get_thread_limit(void);'
614 1.1 mrg
615 1.1 mrg _Fortran_:
616 1.11 mrg _Interface_: 'integer function omp_get_thread_limit()'
617 1.1 mrg
618 1.1 mrg _See also_:
619 1.1 mrg *note omp_get_max_threads::, *note OMP_THREAD_LIMIT::
620 1.1 mrg
621 1.1 mrg _Reference_:
622 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.14.
623 1.1 mrg
624 1.1 mrg
625 1.1 mrg File: libgomp.info, Node: omp_get_thread_num, Next: omp_in_parallel, Prev: omp_get_thread_limit, Up: Runtime Library Routines
626 1.1 mrg
627 1.11 mrg 2.20 'omp_get_thread_num' - Current thread ID
628 1.1 mrg =============================================
629 1.1 mrg
630 1.1 mrg _Description_:
631 1.3 mrg Returns a unique thread identification number within the current
632 1.11 mrg team. In a sequential parts of the program, 'omp_get_thread_num'
633 1.11 mrg always returns 0. In parallel regions the return value varies from
634 1.11 mrg 0 to 'omp_get_num_threads'-1 inclusive. The return value of the
635 1.11 mrg master thread of a team is always 0.
636 1.1 mrg
637 1.1 mrg _C/C++_:
638 1.11 mrg _Prototype_: 'int omp_get_thread_num(void);'
639 1.1 mrg
640 1.1 mrg _Fortran_:
641 1.11 mrg _Interface_: 'integer function omp_get_thread_num()'
642 1.1 mrg
643 1.1 mrg _See also_:
644 1.1 mrg *note omp_get_num_threads::, *note omp_get_ancestor_thread_num::
645 1.1 mrg
646 1.1 mrg _Reference_:
647 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.4.
648 1.1 mrg
649 1.1 mrg
650 1.3 mrg File: libgomp.info, Node: omp_in_parallel, Next: omp_in_final, Prev: omp_get_thread_num, Up: Runtime Library Routines
651 1.1 mrg
652 1.11 mrg 2.21 'omp_in_parallel' - Whether a parallel region is active
653 1.1 mrg ============================================================
654 1.1 mrg
655 1.1 mrg _Description_:
656 1.11 mrg This function returns 'true' if currently running in parallel,
657 1.11 mrg 'false' otherwise. Here, 'true' and 'false' represent their
658 1.1 mrg language-specific counterparts.
659 1.1 mrg
660 1.1 mrg _C/C++_:
661 1.11 mrg _Prototype_: 'int omp_in_parallel(void);'
662 1.1 mrg
663 1.1 mrg _Fortran_:
664 1.11 mrg _Interface_: 'logical function omp_in_parallel()'
665 1.1 mrg
666 1.1 mrg _Reference_:
667 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.6.
668 1.1 mrg
669 1.1 mrg
670 1.5 mrg File: libgomp.info, Node: omp_in_final, Next: omp_is_initial_device, Prev: omp_in_parallel, Up: Runtime Library Routines
671 1.3 mrg
672 1.11 mrg 2.22 'omp_in_final' - Whether in final or included task region
673 1.3 mrg ==============================================================
674 1.3 mrg
675 1.3 mrg _Description_:
676 1.11 mrg This function returns 'true' if currently running in a final or
677 1.11 mrg included task region, 'false' otherwise. Here, 'true' and 'false'
678 1.3 mrg represent their language-specific counterparts.
679 1.3 mrg
680 1.3 mrg _C/C++_:
681 1.11 mrg _Prototype_: 'int omp_in_final(void);'
682 1.1 mrg
683 1.3 mrg _Fortran_:
684 1.11 mrg _Interface_: 'logical function omp_in_final()'
685 1.3 mrg
686 1.3 mrg _Reference_:
687 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.21.
688 1.5 mrg
689 1.5 mrg
690 1.5 mrg File: libgomp.info, Node: omp_is_initial_device, Next: omp_set_default_device, Prev: omp_in_final, Up: Runtime Library Routines
691 1.5 mrg
692 1.11 mrg 2.23 'omp_is_initial_device' - Whether executing on the host device
693 1.5 mrg ===================================================================
694 1.5 mrg
695 1.5 mrg _Description_:
696 1.11 mrg This function returns 'true' if currently running on the host
697 1.11 mrg device, 'false' otherwise. Here, 'true' and 'false' represent
698 1.5 mrg their language-specific counterparts.
699 1.5 mrg
700 1.5 mrg _C/C++_:
701 1.11 mrg _Prototype_: 'int omp_is_initial_device(void);'
702 1.5 mrg
703 1.5 mrg _Fortran_:
704 1.11 mrg _Interface_: 'logical function omp_is_initial_device()'
705 1.5 mrg
706 1.5 mrg _Reference_:
707 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.34.
708 1.5 mrg
709 1.5 mrg
710 1.5 mrg File: libgomp.info, Node: omp_set_default_device, Next: omp_set_dynamic, Prev: omp_is_initial_device, Up: Runtime Library Routines
711 1.5 mrg
712 1.11 mrg 2.24 'omp_set_default_device' - Set the default device for target regions
713 1.5 mrg =========================================================================
714 1.5 mrg
715 1.5 mrg _Description_:
716 1.5 mrg Set the default device for target regions without device clause.
717 1.5 mrg The argument shall be a nonnegative device number.
718 1.5 mrg
719 1.5 mrg _C/C++_:
720 1.11 mrg _Prototype_: 'void omp_set_default_device(int device_num);'
721 1.5 mrg
722 1.5 mrg _Fortran_:
723 1.11 mrg _Interface_: 'subroutine omp_set_default_device(device_num)'
724 1.11 mrg 'integer device_num'
725 1.5 mrg
726 1.5 mrg _See also_:
727 1.5 mrg *note OMP_DEFAULT_DEVICE::, *note omp_get_default_device::
728 1.5 mrg
729 1.5 mrg _Reference_:
730 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.29.
731 1.3 mrg
732 1.3 mrg
733 1.5 mrg File: libgomp.info, Node: omp_set_dynamic, Next: omp_set_max_active_levels, Prev: omp_set_default_device, Up: Runtime Library Routines
734 1.3 mrg
735 1.11 mrg 2.25 'omp_set_dynamic' - Enable/disable dynamic teams
736 1.1 mrg =====================================================
737 1.1 mrg
738 1.1 mrg _Description_:
739 1.1 mrg Enable or disable the dynamic adjustment of the number of threads
740 1.5 mrg within a team. The function takes the language-specific equivalent
741 1.11 mrg of 'true' and 'false', where 'true' enables dynamic adjustment of
742 1.11 mrg team sizes and 'false' disables it.
743 1.1 mrg
744 1.1 mrg _C/C++_:
745 1.11 mrg _Prototype_: 'void omp_set_dynamic(int dynamic_threads);'
746 1.1 mrg
747 1.1 mrg _Fortran_:
748 1.11 mrg _Interface_: 'subroutine omp_set_dynamic(dynamic_threads)'
749 1.11 mrg 'logical, intent(in) :: dynamic_threads'
750 1.1 mrg
751 1.1 mrg _See also_:
752 1.1 mrg *note OMP_DYNAMIC::, *note omp_get_dynamic::
753 1.1 mrg
754 1.1 mrg _Reference_:
755 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.7.
756 1.1 mrg
757 1.1 mrg
758 1.1 mrg File: libgomp.info, Node: omp_set_max_active_levels, Next: omp_set_nested, Prev: omp_set_dynamic, Up: Runtime Library Routines
759 1.1 mrg
760 1.11 mrg 2.26 'omp_set_max_active_levels' - Limits the number of active parallel regions
761 1.1 mrg ===============================================================================
762 1.1 mrg
763 1.1 mrg _Description_:
764 1.3 mrg This function limits the maximum allowed number of nested, active
765 1.3 mrg parallel regions.
766 1.1 mrg
767 1.1 mrg _C/C++_
768 1.11 mrg _Prototype_: 'void omp_set_max_active_levels(int max_levels);'
769 1.1 mrg
770 1.1 mrg _Fortran_:
771 1.11 mrg _Interface_: 'subroutine omp_set_max_active_levels(max_levels)'
772 1.11 mrg 'integer max_levels'
773 1.1 mrg
774 1.1 mrg _See also_:
775 1.1 mrg *note omp_get_max_active_levels::, *note omp_get_active_level::
776 1.1 mrg
777 1.1 mrg _Reference_:
778 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.15.
779 1.1 mrg
780 1.1 mrg
781 1.1 mrg File: libgomp.info, Node: omp_set_nested, Next: omp_set_num_threads, Prev: omp_set_max_active_levels, Up: Runtime Library Routines
782 1.1 mrg
783 1.11 mrg 2.27 'omp_set_nested' - Enable/disable nested parallel regions
784 1.1 mrg ==============================================================
785 1.1 mrg
786 1.1 mrg _Description_:
787 1.1 mrg Enable or disable nested parallel regions, i.e., whether team
788 1.5 mrg members are allowed to create new teams. The function takes the
789 1.11 mrg language-specific equivalent of 'true' and 'false', where 'true'
790 1.11 mrg enables dynamic adjustment of team sizes and 'false' disables it.
791 1.1 mrg
792 1.1 mrg _C/C++_:
793 1.11 mrg _Prototype_: 'void omp_set_nested(int nested);'
794 1.1 mrg
795 1.1 mrg _Fortran_:
796 1.11 mrg _Interface_: 'subroutine omp_set_nested(nested)'
797 1.11 mrg 'logical, intent(in) :: nested'
798 1.1 mrg
799 1.1 mrg _See also_:
800 1.1 mrg *note OMP_NESTED::, *note omp_get_nested::
801 1.1 mrg
802 1.1 mrg _Reference_:
803 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.10.
804 1.1 mrg
805 1.1 mrg
806 1.1 mrg File: libgomp.info, Node: omp_set_num_threads, Next: omp_set_schedule, Prev: omp_set_nested, Up: Runtime Library Routines
807 1.1 mrg
808 1.11 mrg 2.28 'omp_set_num_threads' - Set upper team size limit
809 1.1 mrg ======================================================
810 1.1 mrg
811 1.1 mrg _Description_:
812 1.1 mrg Specifies the number of threads used by default in subsequent
813 1.11 mrg parallel sections, if those do not specify a 'num_threads' clause.
814 1.11 mrg The argument of 'omp_set_num_threads' shall be a positive integer.
815 1.1 mrg
816 1.1 mrg _C/C++_:
817 1.11 mrg _Prototype_: 'void omp_set_num_threads(int num_threads);'
818 1.1 mrg
819 1.1 mrg _Fortran_:
820 1.11 mrg _Interface_: 'subroutine omp_set_num_threads(num_threads)'
821 1.11 mrg 'integer, intent(in) :: num_threads'
822 1.1 mrg
823 1.1 mrg _See also_:
824 1.1 mrg *note OMP_NUM_THREADS::, *note omp_get_num_threads::, *note
825 1.1 mrg omp_get_max_threads::
826 1.1 mrg
827 1.1 mrg _Reference_:
828 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.1.
829 1.1 mrg
830 1.1 mrg
831 1.1 mrg File: libgomp.info, Node: omp_set_schedule, Next: omp_init_lock, Prev: omp_set_num_threads, Up: Runtime Library Routines
832 1.1 mrg
833 1.11 mrg 2.29 'omp_set_schedule' - Set the runtime scheduling method
834 1.1 mrg ===========================================================
835 1.1 mrg
836 1.1 mrg _Description_:
837 1.5 mrg Sets the runtime scheduling method. The KIND argument can have the
838 1.11 mrg value 'omp_sched_static', 'omp_sched_dynamic', 'omp_sched_guided'
839 1.11 mrg or 'omp_sched_auto'. Except for 'omp_sched_auto', the chunk size
840 1.7 mrg is set to the value of CHUNK_SIZE if positive, or to the default
841 1.11 mrg value if zero or negative. For 'omp_sched_auto' the CHUNK_SIZE
842 1.1 mrg argument is ignored.
843 1.1 mrg
844 1.1 mrg _C/C++_
845 1.11 mrg _Prototype_: 'void omp_set_schedule(omp_sched_t kind, int
846 1.11 mrg chunk_size);'
847 1.1 mrg
848 1.1 mrg _Fortran_:
849 1.11 mrg _Interface_: 'subroutine omp_set_schedule(kind, chunk_size)'
850 1.11 mrg 'integer(kind=omp_sched_kind) kind'
851 1.11 mrg 'integer chunk_size'
852 1.1 mrg
853 1.1 mrg _See also_:
854 1.1 mrg *note omp_get_schedule:: *note OMP_SCHEDULE::
855 1.1 mrg
856 1.1 mrg _Reference_:
857 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.12.
858 1.1 mrg
859 1.1 mrg
860 1.1 mrg File: libgomp.info, Node: omp_init_lock, Next: omp_set_lock, Prev: omp_set_schedule, Up: Runtime Library Routines
861 1.1 mrg
862 1.11 mrg 2.30 'omp_init_lock' - Initialize simple lock
863 1.1 mrg =============================================
864 1.1 mrg
865 1.1 mrg _Description_:
866 1.3 mrg Initialize a simple lock. After initialization, the lock is in an
867 1.1 mrg unlocked state.
868 1.1 mrg
869 1.1 mrg _C/C++_:
870 1.11 mrg _Prototype_: 'void omp_init_lock(omp_lock_t *lock);'
871 1.1 mrg
872 1.1 mrg _Fortran_:
873 1.11 mrg _Interface_: 'subroutine omp_init_lock(svar)'
874 1.11 mrg 'integer(omp_lock_kind), intent(out) :: svar'
875 1.1 mrg
876 1.1 mrg _See also_:
877 1.1 mrg *note omp_destroy_lock::
878 1.1 mrg
879 1.1 mrg _Reference_:
880 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.1.
881 1.1 mrg
882 1.1 mrg
883 1.1 mrg File: libgomp.info, Node: omp_set_lock, Next: omp_test_lock, Prev: omp_init_lock, Up: Runtime Library Routines
884 1.1 mrg
885 1.11 mrg 2.31 'omp_set_lock' - Wait for and set simple lock
886 1.1 mrg ==================================================
887 1.1 mrg
888 1.1 mrg _Description_:
889 1.11 mrg Before setting a simple lock, the lock variable must be initialized
890 1.11 mrg by 'omp_init_lock'. The calling thread is blocked until the lock
891 1.11 mrg is available. If the lock is already held by the current thread, a
892 1.11 mrg deadlock occurs.
893 1.1 mrg
894 1.1 mrg _C/C++_:
895 1.11 mrg _Prototype_: 'void omp_set_lock(omp_lock_t *lock);'
896 1.1 mrg
897 1.1 mrg _Fortran_:
898 1.11 mrg _Interface_: 'subroutine omp_set_lock(svar)'
899 1.11 mrg 'integer(omp_lock_kind), intent(inout) :: svar'
900 1.1 mrg
901 1.1 mrg _See also_:
902 1.1 mrg *note omp_init_lock::, *note omp_test_lock::, *note
903 1.1 mrg omp_unset_lock::
904 1.1 mrg
905 1.1 mrg _Reference_:
906 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.4.
907 1.1 mrg
908 1.1 mrg
909 1.1 mrg File: libgomp.info, Node: omp_test_lock, Next: omp_unset_lock, Prev: omp_set_lock, Up: Runtime Library Routines
910 1.1 mrg
911 1.11 mrg 2.32 'omp_test_lock' - Test and set simple lock if available
912 1.1 mrg ============================================================
913 1.1 mrg
914 1.1 mrg _Description_:
915 1.11 mrg Before setting a simple lock, the lock variable must be initialized
916 1.11 mrg by 'omp_init_lock'. Contrary to 'omp_set_lock', 'omp_test_lock'
917 1.11 mrg does not block if the lock is not available. This function returns
918 1.11 mrg 'true' upon success, 'false' otherwise. Here, 'true' and 'false'
919 1.11 mrg represent their language-specific counterparts.
920 1.1 mrg
921 1.1 mrg _C/C++_:
922 1.11 mrg _Prototype_: 'int omp_test_lock(omp_lock_t *lock);'
923 1.1 mrg
924 1.1 mrg _Fortran_:
925 1.11 mrg _Interface_: 'logical function omp_test_lock(svar)'
926 1.11 mrg 'integer(omp_lock_kind), intent(inout) :: svar'
927 1.1 mrg
928 1.1 mrg _See also_:
929 1.1 mrg *note omp_init_lock::, *note omp_set_lock::, *note omp_set_lock::
930 1.1 mrg
931 1.1 mrg _Reference_:
932 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.6.
933 1.1 mrg
934 1.1 mrg
935 1.1 mrg File: libgomp.info, Node: omp_unset_lock, Next: omp_destroy_lock, Prev: omp_test_lock, Up: Runtime Library Routines
936 1.1 mrg
937 1.11 mrg 2.33 'omp_unset_lock' - Unset simple lock
938 1.1 mrg =========================================
939 1.1 mrg
940 1.1 mrg _Description_:
941 1.1 mrg A simple lock about to be unset must have been locked by
942 1.11 mrg 'omp_set_lock' or 'omp_test_lock' before. In addition, the lock
943 1.11 mrg must be held by the thread calling 'omp_unset_lock'. Then, the
944 1.11 mrg lock becomes unlocked. If one or more threads attempted to set the
945 1.11 mrg lock before, one of them is chosen to, again, set the lock to
946 1.1 mrg itself.
947 1.1 mrg
948 1.1 mrg _C/C++_:
949 1.11 mrg _Prototype_: 'void omp_unset_lock(omp_lock_t *lock);'
950 1.1 mrg
951 1.1 mrg _Fortran_:
952 1.11 mrg _Interface_: 'subroutine omp_unset_lock(svar)'
953 1.11 mrg 'integer(omp_lock_kind), intent(inout) :: svar'
954 1.1 mrg
955 1.1 mrg _See also_:
956 1.1 mrg *note omp_set_lock::, *note omp_test_lock::
957 1.1 mrg
958 1.1 mrg _Reference_:
959 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.5.
960 1.1 mrg
961 1.1 mrg
962 1.1 mrg File: libgomp.info, Node: omp_destroy_lock, Next: omp_init_nest_lock, Prev: omp_unset_lock, Up: Runtime Library Routines
963 1.1 mrg
964 1.11 mrg 2.34 'omp_destroy_lock' - Destroy simple lock
965 1.1 mrg =============================================
966 1.1 mrg
967 1.1 mrg _Description_:
968 1.5 mrg Destroy a simple lock. In order to be destroyed, a simple lock
969 1.1 mrg must be in the unlocked state.
970 1.1 mrg
971 1.1 mrg _C/C++_:
972 1.11 mrg _Prototype_: 'void omp_destroy_lock(omp_lock_t *lock);'
973 1.1 mrg
974 1.1 mrg _Fortran_:
975 1.11 mrg _Interface_: 'subroutine omp_destroy_lock(svar)'
976 1.11 mrg 'integer(omp_lock_kind), intent(inout) :: svar'
977 1.1 mrg
978 1.1 mrg _See also_:
979 1.1 mrg *note omp_init_lock::
980 1.1 mrg
981 1.1 mrg _Reference_:
982 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.3.
983 1.1 mrg
984 1.1 mrg
985 1.1 mrg File: libgomp.info, Node: omp_init_nest_lock, Next: omp_set_nest_lock, Prev: omp_destroy_lock, Up: Runtime Library Routines
986 1.1 mrg
987 1.11 mrg 2.35 'omp_init_nest_lock' - Initialize nested lock
988 1.1 mrg ==================================================
989 1.1 mrg
990 1.1 mrg _Description_:
991 1.3 mrg Initialize a nested lock. After initialization, the lock is in an
992 1.1 mrg unlocked state and the nesting count is set to zero.
993 1.1 mrg
994 1.1 mrg _C/C++_:
995 1.11 mrg _Prototype_: 'void omp_init_nest_lock(omp_nest_lock_t *lock);'
996 1.1 mrg
997 1.1 mrg _Fortran_:
998 1.11 mrg _Interface_: 'subroutine omp_init_nest_lock(nvar)'
999 1.11 mrg 'integer(omp_nest_lock_kind), intent(out) :: nvar'
1000 1.1 mrg
1001 1.1 mrg _See also_:
1002 1.1 mrg *note omp_destroy_nest_lock::
1003 1.1 mrg
1004 1.1 mrg _Reference_:
1005 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.1.
1006 1.1 mrg
1007 1.1 mrg
1008 1.1 mrg File: libgomp.info, Node: omp_set_nest_lock, Next: omp_test_nest_lock, Prev: omp_init_nest_lock, Up: Runtime Library Routines
1009 1.1 mrg
1010 1.11 mrg 2.36 'omp_set_nest_lock' - Wait for and set nested lock
1011 1.1 mrg =======================================================
1012 1.1 mrg
1013 1.1 mrg _Description_:
1014 1.11 mrg Before setting a nested lock, the lock variable must be initialized
1015 1.11 mrg by 'omp_init_nest_lock'. The calling thread is blocked until the
1016 1.11 mrg lock is available. If the lock is already held by the current
1017 1.11 mrg thread, the nesting count for the lock is incremented.
1018 1.1 mrg
1019 1.1 mrg _C/C++_:
1020 1.11 mrg _Prototype_: 'void omp_set_nest_lock(omp_nest_lock_t *lock);'
1021 1.1 mrg
1022 1.1 mrg _Fortran_:
1023 1.11 mrg _Interface_: 'subroutine omp_set_nest_lock(nvar)'
1024 1.11 mrg 'integer(omp_nest_lock_kind), intent(inout) :: nvar'
1025 1.1 mrg
1026 1.1 mrg _See also_:
1027 1.1 mrg *note omp_init_nest_lock::, *note omp_unset_nest_lock::
1028 1.1 mrg
1029 1.1 mrg _Reference_:
1030 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.4.
1031 1.1 mrg
1032 1.1 mrg
1033 1.1 mrg File: libgomp.info, Node: omp_test_nest_lock, Next: omp_unset_nest_lock, Prev: omp_set_nest_lock, Up: Runtime Library Routines
1034 1.1 mrg
1035 1.11 mrg 2.37 'omp_test_nest_lock' - Test and set nested lock if available
1036 1.1 mrg =================================================================
1037 1.1 mrg
1038 1.1 mrg _Description_:
1039 1.11 mrg Before setting a nested lock, the lock variable must be initialized
1040 1.11 mrg by 'omp_init_nest_lock'. Contrary to 'omp_set_nest_lock',
1041 1.11 mrg 'omp_test_nest_lock' does not block if the lock is not available.
1042 1.11 mrg If the lock is already held by the current thread, the new nesting
1043 1.11 mrg count is returned. Otherwise, the return value equals zero.
1044 1.1 mrg
1045 1.1 mrg _C/C++_:
1046 1.11 mrg _Prototype_: 'int omp_test_nest_lock(omp_nest_lock_t *lock);'
1047 1.1 mrg
1048 1.1 mrg _Fortran_:
1049 1.11 mrg _Interface_: 'logical function omp_test_nest_lock(nvar)'
1050 1.11 mrg 'integer(omp_nest_lock_kind), intent(inout) :: nvar'
1051 1.1 mrg
1052 1.1 mrg _See also_:
1053 1.1 mrg *note omp_init_lock::, *note omp_set_lock::, *note omp_set_lock::
1054 1.1 mrg
1055 1.1 mrg _Reference_:
1056 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.6.
1057 1.1 mrg
1058 1.1 mrg
1059 1.1 mrg File: libgomp.info, Node: omp_unset_nest_lock, Next: omp_destroy_nest_lock, Prev: omp_test_nest_lock, Up: Runtime Library Routines
1060 1.1 mrg
1061 1.11 mrg 2.38 'omp_unset_nest_lock' - Unset nested lock
1062 1.1 mrg ==============================================
1063 1.1 mrg
1064 1.1 mrg _Description_:
1065 1.1 mrg A nested lock about to be unset must have been locked by
1066 1.11 mrg 'omp_set_nested_lock' or 'omp_test_nested_lock' before. In
1067 1.1 mrg addition, the lock must be held by the thread calling
1068 1.11 mrg 'omp_unset_nested_lock'. If the nesting count drops to zero, the
1069 1.5 mrg lock becomes unlocked. If one ore more threads attempted to set
1070 1.3 mrg the lock before, one of them is chosen to, again, set the lock to
1071 1.1 mrg itself.
1072 1.1 mrg
1073 1.1 mrg _C/C++_:
1074 1.11 mrg _Prototype_: 'void omp_unset_nest_lock(omp_nest_lock_t *lock);'
1075 1.1 mrg
1076 1.1 mrg _Fortran_:
1077 1.11 mrg _Interface_: 'subroutine omp_unset_nest_lock(nvar)'
1078 1.11 mrg 'integer(omp_nest_lock_kind), intent(inout) :: nvar'
1079 1.1 mrg
1080 1.1 mrg _See also_:
1081 1.1 mrg *note omp_set_nest_lock::
1082 1.1 mrg
1083 1.1 mrg _Reference_:
1084 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.5.
1085 1.1 mrg
1086 1.1 mrg
1087 1.1 mrg File: libgomp.info, Node: omp_destroy_nest_lock, Next: omp_get_wtick, Prev: omp_unset_nest_lock, Up: Runtime Library Routines
1088 1.1 mrg
1089 1.11 mrg 2.39 'omp_destroy_nest_lock' - Destroy nested lock
1090 1.1 mrg ==================================================
1091 1.1 mrg
1092 1.1 mrg _Description_:
1093 1.5 mrg Destroy a nested lock. In order to be destroyed, a nested lock
1094 1.1 mrg must be in the unlocked state and its nesting count must equal
1095 1.1 mrg zero.
1096 1.1 mrg
1097 1.1 mrg _C/C++_:
1098 1.11 mrg _Prototype_: 'void omp_destroy_nest_lock(omp_nest_lock_t *);'
1099 1.1 mrg
1100 1.1 mrg _Fortran_:
1101 1.11 mrg _Interface_: 'subroutine omp_destroy_nest_lock(nvar)'
1102 1.11 mrg 'integer(omp_nest_lock_kind), intent(inout) :: nvar'
1103 1.1 mrg
1104 1.1 mrg _See also_:
1105 1.1 mrg *note omp_init_lock::
1106 1.1 mrg
1107 1.1 mrg _Reference_:
1108 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.3.
1109 1.1 mrg
1110 1.1 mrg
1111 1.1 mrg File: libgomp.info, Node: omp_get_wtick, Next: omp_get_wtime, Prev: omp_destroy_nest_lock, Up: Runtime Library Routines
1112 1.1 mrg
1113 1.11 mrg 2.40 'omp_get_wtick' - Get timer precision
1114 1.1 mrg ==========================================
1115 1.1 mrg
1116 1.1 mrg _Description_:
1117 1.1 mrg Gets the timer precision, i.e., the number of seconds between two
1118 1.1 mrg successive clock ticks.
1119 1.1 mrg
1120 1.1 mrg _C/C++_:
1121 1.11 mrg _Prototype_: 'double omp_get_wtick(void);'
1122 1.1 mrg
1123 1.1 mrg _Fortran_:
1124 1.11 mrg _Interface_: 'double precision function omp_get_wtick()'
1125 1.1 mrg
1126 1.1 mrg _See also_:
1127 1.1 mrg *note omp_get_wtime::
1128 1.1 mrg
1129 1.1 mrg _Reference_:
1130 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.4.2.
1131 1.1 mrg
1132 1.1 mrg
1133 1.1 mrg File: libgomp.info, Node: omp_get_wtime, Prev: omp_get_wtick, Up: Runtime Library Routines
1134 1.1 mrg
1135 1.11 mrg 2.41 'omp_get_wtime' - Elapsed wall clock time
1136 1.1 mrg ==============================================
1137 1.1 mrg
1138 1.1 mrg _Description_:
1139 1.5 mrg Elapsed wall clock time in seconds. The time is measured per
1140 1.3 mrg thread, no guarantee can be made that two distinct threads measure
1141 1.3 mrg the same time. Time is measured from some "time in the past",
1142 1.3 mrg which is an arbitrary time guaranteed not to change during the
1143 1.3 mrg execution of the program.
1144 1.1 mrg
1145 1.1 mrg _C/C++_:
1146 1.11 mrg _Prototype_: 'double omp_get_wtime(void);'
1147 1.1 mrg
1148 1.1 mrg _Fortran_:
1149 1.11 mrg _Interface_: 'double precision function omp_get_wtime()'
1150 1.1 mrg
1151 1.1 mrg _See also_:
1152 1.1 mrg *note omp_get_wtick::
1153 1.1 mrg
1154 1.1 mrg _Reference_:
1155 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 3.4.1.
1156 1.1 mrg
1157 1.1 mrg
1158 1.7 mrg File: libgomp.info, Node: Environment Variables, Next: Enabling OpenACC, Prev: Runtime Library Routines, Up: Top
1159 1.1 mrg
1160 1.1 mrg 3 Environment Variables
1161 1.1 mrg ***********************
1162 1.1 mrg
1163 1.11 mrg The environment variables which beginning with 'OMP_' are defined by
1164 1.7 mrg section 4 of the OpenMP specification in version 4.5, while those
1165 1.11 mrg beginning with 'GOMP_' are GNU extensions.
1166 1.1 mrg
1167 1.1 mrg * Menu:
1168 1.1 mrg
1169 1.7 mrg * OMP_CANCELLATION:: Set whether cancellation is activated
1170 1.7 mrg * OMP_DISPLAY_ENV:: Show OpenMP version and environment variables
1171 1.7 mrg * OMP_DEFAULT_DEVICE:: Set the device used in target regions
1172 1.7 mrg * OMP_DYNAMIC:: Dynamic adjustment of threads
1173 1.7 mrg * OMP_MAX_ACTIVE_LEVELS:: Set the maximum number of nested parallel regions
1174 1.7 mrg * OMP_MAX_TASK_PRIORITY:: Set the maximum task priority value
1175 1.7 mrg * OMP_NESTED:: Nested parallel regions
1176 1.7 mrg * OMP_NUM_THREADS:: Specifies the number of threads to use
1177 1.7 mrg * OMP_PROC_BIND:: Whether theads may be moved between CPUs
1178 1.7 mrg * OMP_PLACES:: Specifies on which CPUs the theads should be placed
1179 1.7 mrg * OMP_STACKSIZE:: Set default thread stack size
1180 1.7 mrg * OMP_SCHEDULE:: How threads are scheduled
1181 1.7 mrg * OMP_THREAD_LIMIT:: Set the maximum number of threads
1182 1.7 mrg * OMP_WAIT_POLICY:: How waiting threads are handled
1183 1.7 mrg * GOMP_CPU_AFFINITY:: Bind threads to specific CPUs
1184 1.7 mrg * GOMP_DEBUG:: Enable debugging output
1185 1.7 mrg * GOMP_STACKSIZE:: Set default thread stack size
1186 1.7 mrg * GOMP_SPINCOUNT:: Set the busy-wait spin count
1187 1.7 mrg * GOMP_RTEMS_THREAD_POOLS:: Set the RTEMS specific thread pools
1188 1.1 mrg
1189 1.1 mrg
1190 1.5 mrg File: libgomp.info, Node: OMP_CANCELLATION, Next: OMP_DISPLAY_ENV, Up: Environment Variables
1191 1.5 mrg
1192 1.11 mrg 3.1 'OMP_CANCELLATION' - Set whether cancellation is activated
1193 1.5 mrg ==============================================================
1194 1.5 mrg
1195 1.5 mrg _Description_:
1196 1.11 mrg If set to 'TRUE', the cancellation is activated. If set to 'FALSE'
1197 1.11 mrg or if unset, cancellation is disabled and the 'cancel' construct is
1198 1.11 mrg ignored.
1199 1.5 mrg
1200 1.5 mrg _See also_:
1201 1.5 mrg *note omp_get_cancellation::
1202 1.1 mrg
1203 1.5 mrg _Reference_:
1204 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 4.11
1205 1.5 mrg
1206 1.5 mrg
1207 1.5 mrg File: libgomp.info, Node: OMP_DISPLAY_ENV, Next: OMP_DEFAULT_DEVICE, Prev: OMP_CANCELLATION, Up: Environment Variables
1208 1.5 mrg
1209 1.11 mrg 3.2 'OMP_DISPLAY_ENV' - Show OpenMP version and environment variables
1210 1.5 mrg =====================================================================
1211 1.5 mrg
1212 1.5 mrg _Description_:
1213 1.11 mrg If set to 'TRUE', the OpenMP version number and the values
1214 1.5 mrg associated with the OpenMP environment variables are printed to
1215 1.11 mrg 'stderr'. If set to 'VERBOSE', it additionally shows the value of
1216 1.5 mrg the environment variables which are GNU extensions. If undefined
1217 1.11 mrg or set to 'FALSE', this information will not be shown.
1218 1.5 mrg
1219 1.5 mrg _Reference_:
1220 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 4.12
1221 1.5 mrg
1222 1.5 mrg
1223 1.5 mrg File: libgomp.info, Node: OMP_DEFAULT_DEVICE, Next: OMP_DYNAMIC, Prev: OMP_DISPLAY_ENV, Up: Environment Variables
1224 1.5 mrg
1225 1.11 mrg 3.3 'OMP_DEFAULT_DEVICE' - Set the device used in target regions
1226 1.5 mrg ================================================================
1227 1.5 mrg
1228 1.5 mrg _Description_:
1229 1.11 mrg Set to choose the device which is used in a 'target' region, unless
1230 1.11 mrg the value is overridden by 'omp_set_default_device' or by a
1231 1.11 mrg 'device' clause. The value shall be the nonnegative device number.
1232 1.11 mrg If no device with the given device number exists, the code is
1233 1.11 mrg executed on the host. If unset, device number 0 will be used.
1234 1.5 mrg
1235 1.5 mrg _See also_:
1236 1.5 mrg *note omp_get_default_device::, *note omp_set_default_device::,
1237 1.5 mrg
1238 1.5 mrg _Reference_:
1239 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 4.13
1240 1.5 mrg
1241 1.5 mrg
1242 1.5 mrg File: libgomp.info, Node: OMP_DYNAMIC, Next: OMP_MAX_ACTIVE_LEVELS, Prev: OMP_DEFAULT_DEVICE, Up: Environment Variables
1243 1.5 mrg
1244 1.11 mrg 3.4 'OMP_DYNAMIC' - Dynamic adjustment of threads
1245 1.1 mrg =================================================
1246 1.1 mrg
1247 1.1 mrg _Description_:
1248 1.1 mrg Enable or disable the dynamic adjustment of the number of threads
1249 1.5 mrg within a team. The value of this environment variable shall be
1250 1.11 mrg 'TRUE' or 'FALSE'. If undefined, dynamic adjustment is disabled by
1251 1.11 mrg default.
1252 1.1 mrg
1253 1.1 mrg _See also_:
1254 1.1 mrg *note omp_set_dynamic::
1255 1.1 mrg
1256 1.1 mrg _Reference_:
1257 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 4.3
1258 1.1 mrg
1259 1.1 mrg
1260 1.7 mrg File: libgomp.info, Node: OMP_MAX_ACTIVE_LEVELS, Next: OMP_MAX_TASK_PRIORITY, Prev: OMP_DYNAMIC, Up: Environment Variables
1261 1.1 mrg
1262 1.11 mrg 3.5 'OMP_MAX_ACTIVE_LEVELS' - Set the maximum number of nested parallel regions
1263 1.1 mrg ===============================================================================
1264 1.1 mrg
1265 1.1 mrg _Description_:
1266 1.3 mrg Specifies the initial value for the maximum number of nested
1267 1.5 mrg parallel regions. The value of this variable shall be a positive
1268 1.1 mrg integer. If undefined, the number of active levels is unlimited.
1269 1.1 mrg
1270 1.1 mrg _See also_:
1271 1.1 mrg *note omp_set_max_active_levels::
1272 1.1 mrg
1273 1.1 mrg _Reference_:
1274 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 4.9
1275 1.1 mrg
1276 1.1 mrg
1277 1.7 mrg File: libgomp.info, Node: OMP_MAX_TASK_PRIORITY, Next: OMP_NESTED, Prev: OMP_MAX_ACTIVE_LEVELS, Up: Environment Variables
1278 1.7 mrg
1279 1.11 mrg 3.6 'OMP_MAX_TASK_PRIORITY' - Set the maximum priority
1280 1.7 mrg ======================================================
1281 1.7 mrg
1282 1.11 mrg number that can be set for a task.
1283 1.7 mrg _Description_:
1284 1.11 mrg Specifies the initial value for the maximum priority value that can
1285 1.11 mrg be set for a task. The value of this variable shall be a
1286 1.7 mrg non-negative integer, and zero is allowed. If undefined, the
1287 1.7 mrg default priority is 0.
1288 1.7 mrg
1289 1.7 mrg _See also_:
1290 1.7 mrg *note omp_get_max_task_priority::
1291 1.1 mrg
1292 1.7 mrg _Reference_:
1293 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 4.14
1294 1.7 mrg
1295 1.7 mrg
1296 1.7 mrg File: libgomp.info, Node: OMP_NESTED, Next: OMP_NUM_THREADS, Prev: OMP_MAX_TASK_PRIORITY, Up: Environment Variables
1297 1.7 mrg
1298 1.11 mrg 3.7 'OMP_NESTED' - Nested parallel regions
1299 1.1 mrg ==========================================
1300 1.1 mrg
1301 1.1 mrg _Description_:
1302 1.1 mrg Enable or disable nested parallel regions, i.e., whether team
1303 1.5 mrg members are allowed to create new teams. The value of this
1304 1.11 mrg environment variable shall be 'TRUE' or 'FALSE'. If undefined,
1305 1.1 mrg nested parallel regions are disabled by default.
1306 1.1 mrg
1307 1.1 mrg _See also_:
1308 1.1 mrg *note omp_set_nested::
1309 1.1 mrg
1310 1.1 mrg _Reference_:
1311 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 4.6
1312 1.1 mrg
1313 1.1 mrg
1314 1.5 mrg File: libgomp.info, Node: OMP_NUM_THREADS, Next: OMP_PROC_BIND, Prev: OMP_NESTED, Up: Environment Variables
1315 1.1 mrg
1316 1.11 mrg 3.8 'OMP_NUM_THREADS' - Specifies the number of threads to use
1317 1.1 mrg ==============================================================
1318 1.1 mrg
1319 1.1 mrg _Description_:
1320 1.11 mrg Specifies the default number of threads to use in parallel regions.
1321 1.11 mrg The value of this variable shall be a comma-separated list of
1322 1.11 mrg positive integers; the value specified the number of threads to use
1323 1.11 mrg for the corresponding nested level. If undefined one thread per
1324 1.11 mrg CPU is used.
1325 1.1 mrg
1326 1.1 mrg _See also_:
1327 1.1 mrg *note omp_set_num_threads::
1328 1.1 mrg
1329 1.1 mrg _Reference_:
1330 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 4.2
1331 1.1 mrg
1332 1.1 mrg
1333 1.5 mrg File: libgomp.info, Node: OMP_PROC_BIND, Next: OMP_PLACES, Prev: OMP_NUM_THREADS, Up: Environment Variables
1334 1.5 mrg
1335 1.11 mrg 3.9 'OMP_PROC_BIND' - Whether theads may be moved between CPUs
1336 1.5 mrg ==============================================================
1337 1.5 mrg
1338 1.5 mrg _Description_:
1339 1.5 mrg Specifies whether threads may be moved between processors. If set
1340 1.11 mrg to 'TRUE', OpenMP theads should not be moved; if set to 'FALSE'
1341 1.5 mrg they may be moved. Alternatively, a comma separated list with the
1342 1.11 mrg values 'MASTER', 'CLOSE' and 'SPREAD' can be used to specify the
1343 1.5 mrg thread affinity policy for the corresponding nesting level. With
1344 1.11 mrg 'MASTER' the worker threads are in the same place partition as the
1345 1.11 mrg master thread. With 'CLOSE' those are kept close to the master
1346 1.11 mrg thread in contiguous place partitions. And with 'SPREAD' a sparse
1347 1.5 mrg distribution across the place partitions is used.
1348 1.5 mrg
1349 1.11 mrg When undefined, 'OMP_PROC_BIND' defaults to 'TRUE' when
1350 1.11 mrg 'OMP_PLACES' or 'GOMP_CPU_AFFINITY' is set and 'FALSE' otherwise.
1351 1.5 mrg
1352 1.5 mrg _See also_:
1353 1.5 mrg *note OMP_PLACES::, *note GOMP_CPU_AFFINITY::, *note
1354 1.5 mrg omp_get_proc_bind::
1355 1.5 mrg
1356 1.5 mrg _Reference_:
1357 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 4.4
1358 1.5 mrg
1359 1.5 mrg
1360 1.5 mrg File: libgomp.info, Node: OMP_PLACES, Next: OMP_STACKSIZE, Prev: OMP_PROC_BIND, Up: Environment Variables
1361 1.1 mrg
1362 1.11 mrg 3.10 'OMP_PLACES' - Specifies on which CPUs the theads should be placed
1363 1.7 mrg =======================================================================
1364 1.1 mrg
1365 1.1 mrg _Description_:
1366 1.11 mrg The thread placement can be either specified using an abstract name
1367 1.11 mrg or by an explicit list of the places. The abstract names
1368 1.11 mrg 'threads', 'cores' and 'sockets' can be optionally followed by a
1369 1.5 mrg positive number in parentheses, which denotes the how many places
1370 1.11 mrg shall be created. With 'threads' each place corresponds to a
1371 1.11 mrg single hardware thread; 'cores' to a single core with the
1372 1.11 mrg corresponding number of hardware threads; and with 'sockets' the
1373 1.5 mrg place corresponds to a single socket. The resulting placement can
1374 1.11 mrg be shown by setting the 'OMP_DISPLAY_ENV' environment variable.
1375 1.5 mrg
1376 1.5 mrg Alternatively, the placement can be specified explicitly as
1377 1.5 mrg comma-separated list of places. A place is specified by set of
1378 1.5 mrg nonnegative numbers in curly braces, denoting the denoting the
1379 1.5 mrg hardware threads. The hardware threads belonging to a place can
1380 1.5 mrg either be specified as comma-separated list of nonnegative thread
1381 1.5 mrg numbers or using an interval. Multiple places can also be either
1382 1.5 mrg specified by a comma-separated list of places or by an interval.
1383 1.5 mrg To specify an interval, a colon followed by the count is placed
1384 1.5 mrg after after the hardware thread number or the place. Optionally,
1385 1.5 mrg the length can be followed by a colon and the stride number -
1386 1.5 mrg otherwise a unit stride is assumed. For instance, the following
1387 1.11 mrg specifies the same places list: '"{0,1,2}, {3,4,6}, {7,8,9},
1388 1.11 mrg {10,11,12}"'; '"{0:3}, {3:3}, {7:3}, {10:3}"'; and '"{0:2}:4:3"'.
1389 1.5 mrg
1390 1.11 mrg If 'OMP_PLACES' and 'GOMP_CPU_AFFINITY' are unset and
1391 1.11 mrg 'OMP_PROC_BIND' is either unset or 'false', threads may be moved
1392 1.5 mrg between CPUs following no placement policy.
1393 1.1 mrg
1394 1.1 mrg _See also_:
1395 1.5 mrg *note OMP_PROC_BIND::, *note GOMP_CPU_AFFINITY::, *note
1396 1.5 mrg omp_get_proc_bind::, *note OMP_DISPLAY_ENV::
1397 1.1 mrg
1398 1.1 mrg _Reference_:
1399 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 4.5
1400 1.1 mrg
1401 1.1 mrg
1402 1.5 mrg File: libgomp.info, Node: OMP_STACKSIZE, Next: OMP_SCHEDULE, Prev: OMP_PLACES, Up: Environment Variables
1403 1.1 mrg
1404 1.11 mrg 3.11 'OMP_STACKSIZE' - Set default thread stack size
1405 1.5 mrg ====================================================
1406 1.1 mrg
1407 1.1 mrg _Description_:
1408 1.1 mrg Set the default thread stack size in kilobytes, unless the number
1409 1.11 mrg is suffixed by 'B', 'K', 'M' or 'G', in which case the size is,
1410 1.11 mrg respectively, in bytes, kilobytes, megabytes or gigabytes. This is
1411 1.11 mrg different from 'pthread_attr_setstacksize' which gets the number of
1412 1.11 mrg bytes as an argument. If the stack size cannot be set due to
1413 1.11 mrg system constraints, an error is reported and the initial stack size
1414 1.11 mrg is left unchanged. If undefined, the stack size is system
1415 1.11 mrg dependent.
1416 1.1 mrg
1417 1.1 mrg _Reference_:
1418 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 4.7
1419 1.1 mrg
1420 1.1 mrg
1421 1.5 mrg File: libgomp.info, Node: OMP_SCHEDULE, Next: OMP_THREAD_LIMIT, Prev: OMP_STACKSIZE, Up: Environment Variables
1422 1.1 mrg
1423 1.11 mrg 3.12 'OMP_SCHEDULE' - How threads are scheduled
1424 1.5 mrg ===============================================
1425 1.1 mrg
1426 1.1 mrg _Description_:
1427 1.11 mrg Allows to specify 'schedule type' and 'chunk size'. The value of
1428 1.11 mrg the variable shall have the form: 'type[,chunk]' where 'type' is
1429 1.11 mrg one of 'static', 'dynamic', 'guided' or 'auto' The optional 'chunk'
1430 1.11 mrg size shall be a positive integer. If undefined, dynamic scheduling
1431 1.11 mrg and a chunk size of 1 is used.
1432 1.1 mrg
1433 1.1 mrg _See also_:
1434 1.5 mrg *note omp_set_schedule::
1435 1.1 mrg
1436 1.1 mrg _Reference_:
1437 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Sections
1438 1.7 mrg 2.7.1.1 and 4.1
1439 1.1 mrg
1440 1.1 mrg
1441 1.5 mrg File: libgomp.info, Node: OMP_THREAD_LIMIT, Next: OMP_WAIT_POLICY, Prev: OMP_SCHEDULE, Up: Environment Variables
1442 1.1 mrg
1443 1.11 mrg 3.13 'OMP_THREAD_LIMIT' - Set the maximum number of threads
1444 1.5 mrg ===========================================================
1445 1.1 mrg
1446 1.1 mrg _Description_:
1447 1.5 mrg Specifies the number of threads to use for the whole program. The
1448 1.5 mrg value of this variable shall be a positive integer. If undefined,
1449 1.5 mrg the number of threads is not limited.
1450 1.5 mrg
1451 1.5 mrg _See also_:
1452 1.5 mrg *note OMP_NUM_THREADS::, *note omp_get_thread_limit::
1453 1.1 mrg
1454 1.1 mrg _Reference_:
1455 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 4.10
1456 1.3 mrg
1457 1.3 mrg
1458 1.5 mrg File: libgomp.info, Node: OMP_WAIT_POLICY, Next: GOMP_CPU_AFFINITY, Prev: OMP_THREAD_LIMIT, Up: Environment Variables
1459 1.3 mrg
1460 1.11 mrg 3.14 'OMP_WAIT_POLICY' - How waiting threads are handled
1461 1.5 mrg ========================================================
1462 1.3 mrg
1463 1.3 mrg _Description_:
1464 1.5 mrg Specifies whether waiting threads should be active or passive. If
1465 1.11 mrg the value is 'PASSIVE', waiting threads should not consume CPU
1466 1.11 mrg power while waiting; while the value is 'ACTIVE' specifies that
1467 1.5 mrg they should. If undefined, threads wait actively for a short time
1468 1.5 mrg before waiting passively.
1469 1.3 mrg
1470 1.3 mrg _See also_:
1471 1.5 mrg *note GOMP_SPINCOUNT::
1472 1.3 mrg
1473 1.3 mrg _Reference_:
1474 1.7 mrg OpenMP specification v4.5 (http://www.openmp.org/), Section 4.8
1475 1.1 mrg
1476 1.1 mrg
1477 1.5 mrg File: libgomp.info, Node: GOMP_CPU_AFFINITY, Next: GOMP_DEBUG, Prev: OMP_WAIT_POLICY, Up: Environment Variables
1478 1.1 mrg
1479 1.11 mrg 3.15 'GOMP_CPU_AFFINITY' - Bind threads to specific CPUs
1480 1.3 mrg ========================================================
1481 1.1 mrg
1482 1.1 mrg _Description_:
1483 1.5 mrg Binds threads to specific CPUs. The variable should contain a
1484 1.5 mrg space-separated or comma-separated list of CPUs. This list may
1485 1.3 mrg contain different kinds of entries: either single CPU numbers in
1486 1.3 mrg any order, a range of CPUs (M-N) or a range with some stride
1487 1.11 mrg (M-N:S). CPU numbers are zero based. For example,
1488 1.11 mrg 'GOMP_CPU_AFFINITY="0 3 1-2 4-15:2"' will bind the initial thread
1489 1.3 mrg to CPU 0, the second to CPU 3, the third to CPU 1, the fourth to
1490 1.3 mrg CPU 2, the fifth to CPU 4, the sixth through tenth to CPUs 6, 8,
1491 1.3 mrg 10, 12, and 14 respectively and then start assigning back from the
1492 1.11 mrg beginning of the list. 'GOMP_CPU_AFFINITY=0' binds all threads to
1493 1.3 mrg CPU 0.
1494 1.1 mrg
1495 1.5 mrg There is no libgomp library routine to determine whether a CPU
1496 1.5 mrg affinity specification is in effect. As a workaround,
1497 1.11 mrg language-specific library functions, e.g., 'getenv' in C or
1498 1.11 mrg 'GET_ENVIRONMENT_VARIABLE' in Fortran, may be used to query the
1499 1.11 mrg setting of the 'GOMP_CPU_AFFINITY' environment variable. A defined
1500 1.11 mrg CPU affinity on startup cannot be changed or disabled during the
1501 1.11 mrg runtime of the application.
1502 1.11 mrg
1503 1.11 mrg If both 'GOMP_CPU_AFFINITY' and 'OMP_PROC_BIND' are set,
1504 1.11 mrg 'OMP_PROC_BIND' has a higher precedence. If neither has been set
1505 1.11 mrg and 'OMP_PROC_BIND' is unset, or when 'OMP_PROC_BIND' is set to
1506 1.11 mrg 'FALSE', the host system will handle the assignment of threads to
1507 1.5 mrg CPUs.
1508 1.1 mrg
1509 1.3 mrg _See also_:
1510 1.5 mrg *note OMP_PLACES::, *note OMP_PROC_BIND::
1511 1.5 mrg
1512 1.5 mrg
1513 1.5 mrg File: libgomp.info, Node: GOMP_DEBUG, Next: GOMP_STACKSIZE, Prev: GOMP_CPU_AFFINITY, Up: Environment Variables
1514 1.5 mrg
1515 1.11 mrg 3.16 'GOMP_DEBUG' - Enable debugging output
1516 1.5 mrg ===========================================
1517 1.5 mrg
1518 1.5 mrg _Description_:
1519 1.11 mrg Enable debugging output. The variable should be set to '0'
1520 1.11 mrg (disabled, also the default if not set), or '1' (enabled).
1521 1.5 mrg
1522 1.5 mrg If enabled, some debugging output will be printed during execution.
1523 1.5 mrg This is currently not specified in more detail, and subject to
1524 1.5 mrg change.
1525 1.3 mrg
1526 1.1 mrg
1527 1.5 mrg File: libgomp.info, Node: GOMP_STACKSIZE, Next: GOMP_SPINCOUNT, Prev: GOMP_DEBUG, Up: Environment Variables
1528 1.1 mrg
1529 1.11 mrg 3.17 'GOMP_STACKSIZE' - Set default thread stack size
1530 1.1 mrg =====================================================
1531 1.1 mrg
1532 1.1 mrg _Description_:
1533 1.5 mrg Set the default thread stack size in kilobytes. This is different
1534 1.11 mrg from 'pthread_attr_setstacksize' which gets the number of bytes as
1535 1.5 mrg an argument. If the stack size cannot be set due to system
1536 1.3 mrg constraints, an error is reported and the initial stack size is
1537 1.5 mrg left unchanged. If undefined, the stack size is system dependent.
1538 1.1 mrg
1539 1.1 mrg _See also_:
1540 1.3 mrg *note OMP_STACKSIZE::
1541 1.1 mrg
1542 1.1 mrg _Reference_:
1543 1.1 mrg GCC Patches Mailinglist
1544 1.1 mrg (http://gcc.gnu.org/ml/gcc-patches/2006-06/msg00493.html), GCC
1545 1.1 mrg Patches Mailinglist
1546 1.1 mrg (http://gcc.gnu.org/ml/gcc-patches/2006-06/msg00496.html)
1547 1.1 mrg
1548 1.1 mrg
1549 1.7 mrg File: libgomp.info, Node: GOMP_SPINCOUNT, Next: GOMP_RTEMS_THREAD_POOLS, Prev: GOMP_STACKSIZE, Up: Environment Variables
1550 1.5 mrg
1551 1.11 mrg 3.18 'GOMP_SPINCOUNT' - Set the busy-wait spin count
1552 1.5 mrg ====================================================
1553 1.5 mrg
1554 1.5 mrg _Description_:
1555 1.5 mrg Determines how long a threads waits actively with consuming CPU
1556 1.5 mrg power before waiting passively without consuming CPU power. The
1557 1.11 mrg value may be either 'INFINITE', 'INFINITY' to always wait actively
1558 1.5 mrg or an integer which gives the number of spins of the busy-wait
1559 1.5 mrg loop. The integer may optionally be followed by the following
1560 1.11 mrg suffixes acting as multiplication factors: 'k' (kilo, thousand),
1561 1.11 mrg 'M' (mega, million), 'G' (giga, billion), or 'T' (tera, trillion).
1562 1.11 mrg If undefined, 0 is used when 'OMP_WAIT_POLICY' is 'PASSIVE',
1563 1.11 mrg 300,000 is used when 'OMP_WAIT_POLICY' is undefined and 30 billion
1564 1.11 mrg is used when 'OMP_WAIT_POLICY' is 'ACTIVE'. If there are more
1565 1.11 mrg OpenMP threads than available CPUs, 1000 and 100 spins are used for
1566 1.11 mrg 'OMP_WAIT_POLICY' being 'ACTIVE' or undefined, respectively; unless
1567 1.11 mrg the 'GOMP_SPINCOUNT' is lower or 'OMP_WAIT_POLICY' is 'PASSIVE'.
1568 1.5 mrg
1569 1.5 mrg _See also_:
1570 1.5 mrg *note OMP_WAIT_POLICY::
1571 1.5 mrg
1572 1.5 mrg
1573 1.7 mrg File: libgomp.info, Node: GOMP_RTEMS_THREAD_POOLS, Prev: GOMP_SPINCOUNT, Up: Environment Variables
1574 1.7 mrg
1575 1.11 mrg 3.19 'GOMP_RTEMS_THREAD_POOLS' - Set the RTEMS specific thread pools
1576 1.7 mrg ====================================================================
1577 1.7 mrg
1578 1.7 mrg _Description_:
1579 1.7 mrg This environment variable is only used on the RTEMS real-time
1580 1.7 mrg operating system. It determines the scheduler instance specific
1581 1.11 mrg thread pools. The format for 'GOMP_RTEMS_THREAD_POOLS' is a list
1582 1.11 mrg of optional '<thread-pool-count>[$<priority>]@<scheduler-name>'
1583 1.11 mrg configurations separated by ':' where:
1584 1.11 mrg * '<thread-pool-count>' is the thread pool count for this
1585 1.7 mrg scheduler instance.
1586 1.11 mrg * '$<priority>' is an optional priority for the worker threads
1587 1.11 mrg of a thread pool according to 'pthread_setschedparam'. In
1588 1.7 mrg case a priority value is omitted, then a worker thread will
1589 1.7 mrg inherit the priority of the OpenMP master thread that created
1590 1.7 mrg it. The priority of the worker thread is not changed after
1591 1.7 mrg creation, even if a new OpenMP master thread using the worker
1592 1.7 mrg has a different priority.
1593 1.11 mrg * '@<scheduler-name>' is the scheduler instance name according
1594 1.7 mrg to the RTEMS application configuration.
1595 1.7 mrg In case no thread pool configuration is specified for a scheduler
1596 1.11 mrg instance, then each OpenMP master thread of this scheduler instance
1597 1.11 mrg will use its own dynamically allocated thread pool. To limit the
1598 1.11 mrg worker thread count of the thread pools, each OpenMP master thread
1599 1.11 mrg must call 'omp_set_num_threads'.
1600 1.7 mrg _Example_:
1601 1.11 mrg Lets suppose we have three scheduler instances 'IO', 'WRK0', and
1602 1.11 mrg 'WRK1' with 'GOMP_RTEMS_THREAD_POOLS' set to '"1@WRK0:3$4@WRK1"'.
1603 1.7 mrg Then there are no thread pool restrictions for scheduler instance
1604 1.11 mrg 'IO'. In the scheduler instance 'WRK0' there is one thread pool
1605 1.7 mrg available. Since no priority is specified for this scheduler
1606 1.7 mrg instance, the worker thread inherits the priority of the OpenMP
1607 1.11 mrg master thread that created it. In the scheduler instance 'WRK1'
1608 1.11 mrg there are three thread pools available and their worker threads run
1609 1.11 mrg at priority four.
1610 1.7 mrg
1611 1.7 mrg
1612 1.7 mrg File: libgomp.info, Node: Enabling OpenACC, Next: OpenACC Runtime Library Routines, Prev: Environment Variables, Up: Top
1613 1.7 mrg
1614 1.7 mrg 4 Enabling OpenACC
1615 1.7 mrg ******************
1616 1.7 mrg
1617 1.7 mrg To activate the OpenACC extensions for C/C++ and Fortran, the
1618 1.11 mrg compile-time flag '-fopenacc' must be specified. This enables the
1619 1.11 mrg OpenACC directive '#pragma acc' in C/C++ and '!$accp' directives in free
1620 1.11 mrg form, 'c$acc', '*$acc' and '!$acc' directives in fixed form, '!$'
1621 1.11 mrg conditional compilation sentinels in free form and 'c$', '*$' and '!$'
1622 1.7 mrg sentinels in fixed form, for Fortran. The flag also arranges for
1623 1.7 mrg automatic linking of the OpenACC runtime library (*note OpenACC Runtime
1624 1.7 mrg Library Routines::).
1625 1.7 mrg
1626 1.7 mrg A complete description of all OpenACC directives accepted may be
1627 1.11 mrg found in the OpenACC (https://www.openacc.org) Application Programming
1628 1.7 mrg Interface manual, version 2.0.
1629 1.7 mrg
1630 1.7 mrg Note that this is an experimental feature and subject to change in
1631 1.11 mrg future versions of GCC. See <https://gcc.gnu.org/wiki/OpenACC> for more
1632 1.11 mrg information.
1633 1.7 mrg
1634 1.7 mrg
1635 1.7 mrg File: libgomp.info, Node: OpenACC Runtime Library Routines, Next: OpenACC Environment Variables, Prev: Enabling OpenACC, Up: Top
1636 1.7 mrg
1637 1.7 mrg 5 OpenACC Runtime Library Routines
1638 1.7 mrg **********************************
1639 1.7 mrg
1640 1.7 mrg The runtime routines described here are defined by section 3 of the
1641 1.7 mrg OpenACC specifications in version 2.0. They have C linkage, and do not
1642 1.11 mrg throw exceptions. Generally, they are available only for the host, with
1643 1.11 mrg the exception of 'acc_on_device', which is available for both the host
1644 1.11 mrg and the acceleration device.
1645 1.7 mrg
1646 1.7 mrg * Menu:
1647 1.7 mrg
1648 1.7 mrg * acc_get_num_devices:: Get number of devices for the given device
1649 1.7 mrg type.
1650 1.7 mrg * acc_set_device_type:: Set type of device accelerator to use.
1651 1.7 mrg * acc_get_device_type:: Get type of device accelerator to be used.
1652 1.7 mrg * acc_set_device_num:: Set device number to use.
1653 1.7 mrg * acc_get_device_num:: Get device number to be used.
1654 1.7 mrg * acc_async_test:: Tests for completion of a specific asynchronous
1655 1.7 mrg operation.
1656 1.7 mrg * acc_async_test_all:: Tests for completion of all asychronous
1657 1.7 mrg operations.
1658 1.7 mrg * acc_wait:: Wait for completion of a specific asynchronous
1659 1.7 mrg operation.
1660 1.7 mrg * acc_wait_all:: Waits for completion of all asyncrhonous
1661 1.7 mrg operations.
1662 1.7 mrg * acc_wait_all_async:: Wait for completion of all asynchronous
1663 1.7 mrg operations.
1664 1.7 mrg * acc_wait_async:: Wait for completion of asynchronous operations.
1665 1.7 mrg * acc_init:: Initialize runtime for a specific device type.
1666 1.7 mrg * acc_shutdown:: Shuts down the runtime for a specific device
1667 1.7 mrg type.
1668 1.7 mrg * acc_on_device:: Whether executing on a particular device
1669 1.7 mrg * acc_malloc:: Allocate device memory.
1670 1.7 mrg * acc_free:: Free device memory.
1671 1.7 mrg * acc_copyin:: Allocate device memory and copy host memory to
1672 1.7 mrg it.
1673 1.7 mrg * acc_present_or_copyin:: If the data is not present on the device,
1674 1.7 mrg allocate device memory and copy from host
1675 1.7 mrg memory.
1676 1.7 mrg * acc_create:: Allocate device memory and map it to host
1677 1.7 mrg memory.
1678 1.7 mrg * acc_present_or_create:: If the data is not present on the device,
1679 1.7 mrg allocate device memory and map it to host
1680 1.7 mrg memory.
1681 1.7 mrg * acc_copyout:: Copy device memory to host memory.
1682 1.7 mrg * acc_delete:: Free device memory.
1683 1.7 mrg * acc_update_device:: Update device memory from mapped host memory.
1684 1.7 mrg * acc_update_self:: Update host memory from mapped device memory.
1685 1.7 mrg * acc_map_data:: Map previously allocated device memory to host
1686 1.7 mrg memory.
1687 1.7 mrg * acc_unmap_data:: Unmap device memory from host memory.
1688 1.7 mrg * acc_deviceptr:: Get device pointer associated with specific
1689 1.7 mrg host address.
1690 1.7 mrg * acc_hostptr:: Get host pointer associated with specific
1691 1.7 mrg device address.
1692 1.7 mrg * acc_is_present:: Indiciate whether host variable / array is
1693 1.7 mrg present on device.
1694 1.7 mrg * acc_memcpy_to_device:: Copy host memory to device memory.
1695 1.7 mrg * acc_memcpy_from_device:: Copy device memory to host memory.
1696 1.7 mrg
1697 1.7 mrg API routines for target platforms.
1698 1.7 mrg
1699 1.7 mrg * acc_get_current_cuda_device:: Get CUDA device handle.
1700 1.7 mrg * acc_get_current_cuda_context::Get CUDA context handle.
1701 1.7 mrg * acc_get_cuda_stream:: Get CUDA stream handle.
1702 1.7 mrg * acc_set_cuda_stream:: Set CUDA stream handle.
1703 1.7 mrg
1704 1.7 mrg
1705 1.7 mrg File: libgomp.info, Node: acc_get_num_devices, Next: acc_set_device_type, Up: OpenACC Runtime Library Routines
1706 1.7 mrg
1707 1.11 mrg 5.1 'acc_get_num_devices' - Get number of devices for given device type
1708 1.7 mrg =======================================================================
1709 1.7 mrg
1710 1.7 mrg _Description_
1711 1.7 mrg This function returns a value indicating the number of devices
1712 1.7 mrg available for the device type specified in DEVICETYPE.
1713 1.7 mrg
1714 1.7 mrg _C/C++_:
1715 1.11 mrg _Prototype_: 'int acc_get_num_devices(acc_device_t devicetype);'
1716 1.7 mrg
1717 1.7 mrg _Fortran_:
1718 1.11 mrg _Interface_: 'integer function acc_get_num_devices(devicetype)'
1719 1.11 mrg 'integer(kind=acc_device_kind) devicetype'
1720 1.7 mrg
1721 1.7 mrg _Reference_:
1722 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
1723 1.7 mrg 3.2.1.
1724 1.7 mrg
1725 1.7 mrg
1726 1.7 mrg File: libgomp.info, Node: acc_set_device_type, Next: acc_get_device_type, Prev: acc_get_num_devices, Up: OpenACC Runtime Library Routines
1727 1.7 mrg
1728 1.11 mrg 5.2 'acc_set_device_type' - Set type of device accelerator to use.
1729 1.7 mrg ==================================================================
1730 1.7 mrg
1731 1.7 mrg _Description_
1732 1.7 mrg This function indicates to the runtime library which device typr,
1733 1.7 mrg specified in DEVICETYPE, to use when executing a parallel or
1734 1.7 mrg kernels region.
1735 1.7 mrg
1736 1.7 mrg _C/C++_:
1737 1.11 mrg _Prototype_: 'acc_set_device_type(acc_device_t devicetype);'
1738 1.7 mrg
1739 1.7 mrg _Fortran_:
1740 1.11 mrg _Interface_: 'subroutine acc_set_device_type(devicetype)'
1741 1.11 mrg 'integer(kind=acc_device_kind) devicetype'
1742 1.7 mrg
1743 1.7 mrg _Reference_:
1744 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
1745 1.7 mrg 3.2.2.
1746 1.7 mrg
1747 1.7 mrg
1748 1.7 mrg File: libgomp.info, Node: acc_get_device_type, Next: acc_set_device_num, Prev: acc_set_device_type, Up: OpenACC Runtime Library Routines
1749 1.7 mrg
1750 1.11 mrg 5.3 'acc_get_device_type' - Get type of device accelerator to be used.
1751 1.7 mrg ======================================================================
1752 1.7 mrg
1753 1.7 mrg _Description_
1754 1.7 mrg This function returns what device type will be used when executing
1755 1.7 mrg a parallel or kernels region.
1756 1.7 mrg
1757 1.7 mrg _C/C++_:
1758 1.11 mrg _Prototype_: 'acc_device_t acc_get_device_type(void);'
1759 1.7 mrg
1760 1.7 mrg _Fortran_:
1761 1.11 mrg _Interface_: 'function acc_get_device_type(void)'
1762 1.11 mrg 'integer(kind=acc_device_kind) acc_get_device_type'
1763 1.7 mrg
1764 1.7 mrg _Reference_:
1765 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
1766 1.7 mrg 3.2.3.
1767 1.7 mrg
1768 1.7 mrg
1769 1.7 mrg File: libgomp.info, Node: acc_set_device_num, Next: acc_get_device_num, Prev: acc_get_device_type, Up: OpenACC Runtime Library Routines
1770 1.7 mrg
1771 1.11 mrg 5.4 'acc_set_device_num' - Set device number to use.
1772 1.7 mrg ====================================================
1773 1.7 mrg
1774 1.7 mrg _Description_
1775 1.7 mrg This function will indicate to the runtime which device number,
1776 1.7 mrg specified by NUM, associated with the specifed device type
1777 1.7 mrg DEVICETYPE.
1778 1.7 mrg
1779 1.7 mrg _C/C++_:
1780 1.11 mrg _Prototype_: 'acc_set_device_num(int num, acc_device_t devicetype);'
1781 1.7 mrg
1782 1.7 mrg _Fortran_:
1783 1.11 mrg _Interface_: 'subroutine acc_set_device_num(devicenum, devicetype)'
1784 1.11 mrg 'integer devicenum'
1785 1.11 mrg 'integer(kind=acc_device_kind) devicetype'
1786 1.7 mrg
1787 1.7 mrg _Reference_:
1788 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
1789 1.7 mrg 3.2.4.
1790 1.7 mrg
1791 1.7 mrg
1792 1.7 mrg File: libgomp.info, Node: acc_get_device_num, Next: acc_async_test, Prev: acc_set_device_num, Up: OpenACC Runtime Library Routines
1793 1.7 mrg
1794 1.11 mrg 5.5 'acc_get_device_num' - Get device number to be used.
1795 1.7 mrg ========================================================
1796 1.7 mrg
1797 1.7 mrg _Description_
1798 1.7 mrg This function returns which device number associated with the
1799 1.7 mrg specified device type DEVICETYPE, will be used when executing a
1800 1.7 mrg parallel or kernels region.
1801 1.7 mrg
1802 1.7 mrg _C/C++_:
1803 1.11 mrg _Prototype_: 'int acc_get_device_num(acc_device_t devicetype);'
1804 1.7 mrg
1805 1.7 mrg _Fortran_:
1806 1.11 mrg _Interface_: 'function acc_get_device_num(devicetype)'
1807 1.11 mrg 'integer(kind=acc_device_kind) devicetype'
1808 1.11 mrg 'integer acc_get_device_num'
1809 1.7 mrg
1810 1.7 mrg _Reference_:
1811 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
1812 1.7 mrg 3.2.5.
1813 1.7 mrg
1814 1.7 mrg
1815 1.7 mrg File: libgomp.info, Node: acc_async_test, Next: acc_async_test_all, Prev: acc_get_device_num, Up: OpenACC Runtime Library Routines
1816 1.7 mrg
1817 1.11 mrg 5.6 'acc_async_test' - Test for completion of a specific asynchronous operation.
1818 1.7 mrg ================================================================================
1819 1.7 mrg
1820 1.7 mrg _Description_
1821 1.7 mrg This function tests for completion of the asynchrounous operation
1822 1.11 mrg specified in ARG. In C/C++, a non-zero value will be returned to
1823 1.11 mrg indicate the specified asynchronous operation has completed. While
1824 1.11 mrg Fortran will return a 'true'. If the asynchrounous operation has
1825 1.11 mrg not completed, C/C++ returns a zero and Fortran returns a 'false'.
1826 1.7 mrg
1827 1.7 mrg _C/C++_:
1828 1.11 mrg _Prototype_: 'int acc_async_test(int arg);'
1829 1.7 mrg
1830 1.7 mrg _Fortran_:
1831 1.11 mrg _Interface_: 'function acc_async_test(arg)'
1832 1.11 mrg 'integer(kind=acc_handle_kind) arg'
1833 1.11 mrg 'logical acc_async_test'
1834 1.7 mrg
1835 1.7 mrg _Reference_:
1836 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
1837 1.7 mrg 3.2.6.
1838 1.7 mrg
1839 1.7 mrg
1840 1.7 mrg File: libgomp.info, Node: acc_async_test_all, Next: acc_wait, Prev: acc_async_test, Up: OpenACC Runtime Library Routines
1841 1.7 mrg
1842 1.11 mrg 5.7 'acc_async_test_all' - Tests for completion of all asynchronous operations.
1843 1.7 mrg ===============================================================================
1844 1.7 mrg
1845 1.7 mrg _Description_
1846 1.7 mrg This function tests for completion of all asynchrounous operations.
1847 1.7 mrg In C/C++, a non-zero value will be returned to indicate all
1848 1.11 mrg asynchronous operations have completed. While Fortran will return
1849 1.11 mrg a 'true'. If any asynchronous operation has not completed, C/C++
1850 1.11 mrg returns a zero and Fortran returns a 'false'.
1851 1.7 mrg
1852 1.7 mrg _C/C++_:
1853 1.11 mrg _Prototype_: 'int acc_async_test_all(void);'
1854 1.7 mrg
1855 1.7 mrg _Fortran_:
1856 1.11 mrg _Interface_: 'function acc_async_test()'
1857 1.11 mrg 'logical acc_get_device_num'
1858 1.7 mrg
1859 1.7 mrg _Reference_:
1860 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
1861 1.7 mrg 3.2.7.
1862 1.7 mrg
1863 1.7 mrg
1864 1.7 mrg File: libgomp.info, Node: acc_wait, Next: acc_wait_all, Prev: acc_async_test_all, Up: OpenACC Runtime Library Routines
1865 1.7 mrg
1866 1.11 mrg 5.8 'acc_wait' - Wait for completion of a specific asynchronous operation.
1867 1.7 mrg ==========================================================================
1868 1.7 mrg
1869 1.7 mrg _Description_
1870 1.7 mrg This function waits for completion of the asynchronous operation
1871 1.7 mrg specified in ARG.
1872 1.7 mrg
1873 1.7 mrg _C/C++_:
1874 1.11 mrg _Prototype_: 'acc_wait(arg);'
1875 1.11 mrg _Prototype 'acc_async_wait(arg);'
1876 1.11 mrg (OpenACC 1.0
1877 1.11 mrg compatibility)_:
1878 1.7 mrg
1879 1.7 mrg _Fortran_:
1880 1.11 mrg _Interface_: 'subroutine acc_wait(arg)'
1881 1.11 mrg 'integer(acc_handle_kind) arg'
1882 1.11 mrg _Interface 'subroutine acc_async_wait(arg)'
1883 1.11 mrg (OpenACC 1.0
1884 1.11 mrg compatibility)_:
1885 1.11 mrg 'integer(acc_handle_kind) arg'
1886 1.7 mrg
1887 1.7 mrg _Reference_:
1888 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
1889 1.7 mrg 3.2.8.
1890 1.7 mrg
1891 1.7 mrg
1892 1.7 mrg File: libgomp.info, Node: acc_wait_all, Next: acc_wait_all_async, Prev: acc_wait, Up: OpenACC Runtime Library Routines
1893 1.7 mrg
1894 1.11 mrg 5.9 'acc_wait_all' - Waits for completion of all asynchronous operations.
1895 1.7 mrg =========================================================================
1896 1.7 mrg
1897 1.7 mrg _Description_
1898 1.7 mrg This function waits for the completion of all asynchronous
1899 1.7 mrg operations.
1900 1.7 mrg
1901 1.7 mrg _C/C++_:
1902 1.11 mrg _Prototype_: 'acc_wait_all(void);'
1903 1.11 mrg _Prototype 'acc_async_wait_all(void);'
1904 1.11 mrg (OpenACC 1.0
1905 1.11 mrg compatibility)_:
1906 1.7 mrg
1907 1.7 mrg _Fortran_:
1908 1.11 mrg _Interface_: 'subroutine acc_wait_all()'
1909 1.11 mrg _Interface 'subroutine acc_async_wait_all()'
1910 1.11 mrg (OpenACC 1.0
1911 1.11 mrg compatibility)_:
1912 1.7 mrg
1913 1.7 mrg _Reference_:
1914 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
1915 1.7 mrg 3.2.10.
1916 1.7 mrg
1917 1.7 mrg
1918 1.7 mrg File: libgomp.info, Node: acc_wait_all_async, Next: acc_wait_async, Prev: acc_wait_all, Up: OpenACC Runtime Library Routines
1919 1.7 mrg
1920 1.11 mrg 5.10 'acc_wait_all_async' - Wait for completion of all asynchronous operations.
1921 1.7 mrg ===============================================================================
1922 1.7 mrg
1923 1.7 mrg _Description_
1924 1.7 mrg This function enqueues a wait operation on the queue ASYNC for any
1925 1.7 mrg and all asynchronous operations that have been previously enqueued
1926 1.7 mrg on any queue.
1927 1.7 mrg
1928 1.7 mrg _C/C++_:
1929 1.11 mrg _Prototype_: 'acc_wait_all_async(int async);'
1930 1.7 mrg
1931 1.7 mrg _Fortran_:
1932 1.11 mrg _Interface_: 'subroutine acc_wait_all_async(async)'
1933 1.11 mrg 'integer(acc_handle_kind) async'
1934 1.7 mrg
1935 1.7 mrg _Reference_:
1936 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
1937 1.7 mrg 3.2.11.
1938 1.7 mrg
1939 1.7 mrg
1940 1.7 mrg File: libgomp.info, Node: acc_wait_async, Next: acc_init, Prev: acc_wait_all_async, Up: OpenACC Runtime Library Routines
1941 1.7 mrg
1942 1.11 mrg 5.11 'acc_wait_async' - Wait for completion of asynchronous operations.
1943 1.7 mrg =======================================================================
1944 1.7 mrg
1945 1.7 mrg _Description_
1946 1.7 mrg This function enqueues a wait operation on queue ASYNC for any and
1947 1.7 mrg all asynchronous operations enqueued on queue ARG.
1948 1.7 mrg
1949 1.7 mrg _C/C++_:
1950 1.11 mrg _Prototype_: 'acc_wait_async(int arg, int async);'
1951 1.7 mrg
1952 1.7 mrg _Fortran_:
1953 1.11 mrg _Interface_: 'subroutine acc_wait_async(arg, async)'
1954 1.11 mrg 'integer(acc_handle_kind) arg, async'
1955 1.7 mrg
1956 1.7 mrg _Reference_:
1957 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
1958 1.7 mrg 3.2.9.
1959 1.7 mrg
1960 1.7 mrg
1961 1.7 mrg File: libgomp.info, Node: acc_init, Next: acc_shutdown, Prev: acc_wait_async, Up: OpenACC Runtime Library Routines
1962 1.7 mrg
1963 1.11 mrg 5.12 'acc_init' - Initialize runtime for a specific device type.
1964 1.7 mrg ================================================================
1965 1.7 mrg
1966 1.7 mrg _Description_
1967 1.11 mrg This function initializes the runtime for the device type specified
1968 1.11 mrg in DEVICETYPE.
1969 1.7 mrg
1970 1.7 mrg _C/C++_:
1971 1.11 mrg _Prototype_: 'acc_init(acc_device_t devicetype);'
1972 1.7 mrg
1973 1.7 mrg _Fortran_:
1974 1.11 mrg _Interface_: 'subroutine acc_init(devicetype)'
1975 1.11 mrg 'integer(acc_device_kind) devicetype'
1976 1.7 mrg
1977 1.7 mrg _Reference_:
1978 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
1979 1.7 mrg 3.2.12.
1980 1.7 mrg
1981 1.7 mrg
1982 1.7 mrg File: libgomp.info, Node: acc_shutdown, Next: acc_on_device, Prev: acc_init, Up: OpenACC Runtime Library Routines
1983 1.7 mrg
1984 1.11 mrg 5.13 'acc_shutdown' - Shuts down the runtime for a specific device type.
1985 1.7 mrg ========================================================================
1986 1.7 mrg
1987 1.7 mrg _Description_
1988 1.7 mrg This function shuts down the runtime for the device type specified
1989 1.7 mrg in DEVICETYPE.
1990 1.7 mrg
1991 1.7 mrg _C/C++_:
1992 1.11 mrg _Prototype_: 'acc_shutdown(acc_device_t devicetype);'
1993 1.7 mrg
1994 1.7 mrg _Fortran_:
1995 1.11 mrg _Interface_: 'subroutine acc_shutdown(devicetype)'
1996 1.11 mrg 'integer(acc_device_kind) devicetype'
1997 1.7 mrg
1998 1.7 mrg _Reference_:
1999 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2000 1.7 mrg 3.2.13.
2001 1.7 mrg
2002 1.7 mrg
2003 1.7 mrg File: libgomp.info, Node: acc_on_device, Next: acc_malloc, Prev: acc_shutdown, Up: OpenACC Runtime Library Routines
2004 1.7 mrg
2005 1.11 mrg 5.14 'acc_on_device' - Whether executing on a particular device
2006 1.7 mrg ===============================================================
2007 1.7 mrg
2008 1.7 mrg _Description_:
2009 1.7 mrg This function returns whether the program is executing on a
2010 1.11 mrg particular device specified in DEVICETYPE. In C/C++ a non-zero
2011 1.7 mrg value is returned to indicate the device is execiting on the
2012 1.11 mrg specified device type. In Fortran, 'true' will be returned. If
2013 1.7 mrg the program is not executing on the specified device type C/C++
2014 1.11 mrg will return a zero, while Fortran will return 'false'.
2015 1.7 mrg
2016 1.7 mrg _C/C++_:
2017 1.11 mrg _Prototype_: 'acc_on_device(acc_device_t devicetype);'
2018 1.7 mrg
2019 1.7 mrg _Fortran_:
2020 1.11 mrg _Interface_: 'function acc_on_device(devicetype)'
2021 1.11 mrg 'integer(acc_device_kind) devicetype'
2022 1.11 mrg 'logical acc_on_device'
2023 1.7 mrg
2024 1.7 mrg _Reference_:
2025 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2026 1.7 mrg 3.2.14.
2027 1.7 mrg
2028 1.7 mrg
2029 1.7 mrg File: libgomp.info, Node: acc_malloc, Next: acc_free, Prev: acc_on_device, Up: OpenACC Runtime Library Routines
2030 1.7 mrg
2031 1.11 mrg 5.15 'acc_malloc' - Allocate device memory.
2032 1.7 mrg ===========================================
2033 1.7 mrg
2034 1.7 mrg _Description_
2035 1.11 mrg This function allocates LEN bytes of device memory. It returns the
2036 1.7 mrg device address of the allocated memory.
2037 1.7 mrg
2038 1.7 mrg _C/C++_:
2039 1.11 mrg _Prototype_: 'd_void* acc_malloc(size_t len);'
2040 1.7 mrg
2041 1.7 mrg _Reference_:
2042 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2043 1.7 mrg 3.2.15.
2044 1.7 mrg
2045 1.7 mrg
2046 1.7 mrg File: libgomp.info, Node: acc_free, Next: acc_copyin, Prev: acc_malloc, Up: OpenACC Runtime Library Routines
2047 1.7 mrg
2048 1.11 mrg 5.16 'acc_free' - Free device memory.
2049 1.7 mrg =====================================
2050 1.7 mrg
2051 1.7 mrg _Description_
2052 1.11 mrg Free previously allocated device memory at the device address 'a'.
2053 1.7 mrg
2054 1.7 mrg _C/C++_:
2055 1.11 mrg _Prototype_: 'acc_free(d_void *a);'
2056 1.7 mrg
2057 1.7 mrg _Reference_:
2058 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2059 1.7 mrg 3.2.16.
2060 1.7 mrg
2061 1.7 mrg
2062 1.7 mrg File: libgomp.info, Node: acc_copyin, Next: acc_present_or_copyin, Prev: acc_free, Up: OpenACC Runtime Library Routines
2063 1.7 mrg
2064 1.11 mrg 5.17 'acc_copyin' - Allocate device memory and copy host memory to it.
2065 1.7 mrg ======================================================================
2066 1.7 mrg
2067 1.7 mrg _Description_
2068 1.7 mrg In C/C++, this function allocates LEN bytes of device memory and
2069 1.11 mrg maps it to the specified host address in A. The device address of
2070 1.7 mrg the newly allocated device memory is returned.
2071 1.7 mrg
2072 1.11 mrg In Fortran, two (2) forms are supported. In the first form, A
2073 1.11 mrg specifies a contiguous array section. The second form A specifies
2074 1.11 mrg a variable or array element and LEN specifies the length in bytes.
2075 1.7 mrg
2076 1.7 mrg _C/C++_:
2077 1.11 mrg _Prototype_: 'void *acc_copyin(h_void *a, size_t len);'
2078 1.7 mrg
2079 1.7 mrg _Fortran_:
2080 1.11 mrg _Interface_: 'subroutine acc_copyin(a)'
2081 1.11 mrg 'type, dimension(:[,:]...) :: a'
2082 1.11 mrg _Interface_: 'subroutine acc_copyin(a, len)'
2083 1.11 mrg 'type, dimension(:[,:]...) :: a'
2084 1.11 mrg 'integer len'
2085 1.7 mrg
2086 1.7 mrg _Reference_:
2087 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2088 1.7 mrg 3.2.17.
2089 1.7 mrg
2090 1.7 mrg
2091 1.7 mrg File: libgomp.info, Node: acc_present_or_copyin, Next: acc_create, Prev: acc_copyin, Up: OpenACC Runtime Library Routines
2092 1.7 mrg
2093 1.11 mrg 5.18 'acc_present_or_copyin' - If the data is not present on the device, allocate device memory and copy from host memory.
2094 1.7 mrg ==========================================================================================================================
2095 1.7 mrg
2096 1.7 mrg _Description_
2097 1.7 mrg This function tests if the host data specifed by A and of length
2098 1.11 mrg LEN is present or not. If it is not present, then device memory
2099 1.11 mrg will be allocated and the host memory copied. The device address
2100 1.11 mrg of the newly allocated device memory is returned.
2101 1.7 mrg
2102 1.11 mrg In Fortran, two (2) forms are supported. In the first form, A
2103 1.11 mrg specifies a contiguous array section. The second form A specifies
2104 1.7 mrg a variable or array element and LEN specifies the length in bytes.
2105 1.7 mrg
2106 1.7 mrg _C/C++_:
2107 1.11 mrg _Prototype_: 'void *acc_present_or_copyin(h_void *a, size_t len);'
2108 1.11 mrg _Prototype_: 'void *acc_pcopyin(h_void *a, size_t len);'
2109 1.7 mrg
2110 1.7 mrg _Fortran_:
2111 1.11 mrg _Interface_: 'subroutine acc_present_or_copyin(a)'
2112 1.11 mrg 'type, dimension(:[,:]...) :: a'
2113 1.11 mrg _Interface_: 'subroutine acc_present_or_copyin(a, len)'
2114 1.11 mrg 'type, dimension(:[,:]...) :: a'
2115 1.11 mrg 'integer len'
2116 1.11 mrg _Interface_: 'subroutine acc_pcopyin(a)'
2117 1.11 mrg 'type, dimension(:[,:]...) :: a'
2118 1.11 mrg _Interface_: 'subroutine acc_pcopyin(a, len)'
2119 1.11 mrg 'type, dimension(:[,:]...) :: a'
2120 1.11 mrg 'integer len'
2121 1.7 mrg
2122 1.7 mrg _Reference_:
2123 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2124 1.7 mrg 3.2.18.
2125 1.7 mrg
2126 1.7 mrg
2127 1.7 mrg File: libgomp.info, Node: acc_create, Next: acc_present_or_create, Prev: acc_present_or_copyin, Up: OpenACC Runtime Library Routines
2128 1.7 mrg
2129 1.11 mrg 5.19 'acc_create' - Allocate device memory and map it to host memory.
2130 1.7 mrg =====================================================================
2131 1.7 mrg
2132 1.7 mrg _Description_
2133 1.7 mrg This function allocates device memory and maps it to host memory
2134 1.11 mrg specified by the host address A with a length of LEN bytes. In
2135 1.7 mrg C/C++, the function returns the device address of the allocated
2136 1.7 mrg device memory.
2137 1.7 mrg
2138 1.11 mrg In Fortran, two (2) forms are supported. In the first form, A
2139 1.11 mrg specifies a contiguous array section. The second form A specifies
2140 1.7 mrg a variable or array element and LEN specifies the length in bytes.
2141 1.7 mrg
2142 1.7 mrg _C/C++_:
2143 1.11 mrg _Prototype_: 'void *acc_create(h_void *a, size_t len);'
2144 1.7 mrg
2145 1.7 mrg _Fortran_:
2146 1.11 mrg _Interface_: 'subroutine acc_create(a)'
2147 1.11 mrg 'type, dimension(:[,:]...) :: a'
2148 1.11 mrg _Interface_: 'subroutine acc_create(a, len)'
2149 1.11 mrg 'type, dimension(:[,:]...) :: a'
2150 1.11 mrg 'integer len'
2151 1.7 mrg
2152 1.7 mrg _Reference_:
2153 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2154 1.7 mrg 3.2.19.
2155 1.7 mrg
2156 1.7 mrg
2157 1.7 mrg File: libgomp.info, Node: acc_present_or_create, Next: acc_copyout, Prev: acc_create, Up: OpenACC Runtime Library Routines
2158 1.7 mrg
2159 1.11 mrg 5.20 'acc_present_or_create' - If the data is not present on the device, allocate device memory and map it to host memory.
2160 1.7 mrg ==========================================================================================================================
2161 1.1 mrg
2162 1.7 mrg _Description_
2163 1.7 mrg This function tests if the host data specifed by A and of length
2164 1.11 mrg LEN is present or not. If it is not present, then device memory
2165 1.11 mrg will be allocated and mapped to host memory. In C/C++, the device
2166 1.7 mrg address of the newly allocated device memory is returned.
2167 1.7 mrg
2168 1.11 mrg In Fortran, two (2) forms are supported. In the first form, A
2169 1.11 mrg specifies a contiguous array section. The second form A specifies
2170 1.7 mrg a variable or array element and LEN specifies the length in bytes.
2171 1.7 mrg
2172 1.7 mrg _C/C++_:
2173 1.11 mrg _Prototype_: 'void *acc_present_or_create(h_void *a, size_t len)'
2174 1.11 mrg _Prototype_: 'void *acc_pcreate(h_void *a, size_t len)'
2175 1.7 mrg
2176 1.7 mrg _Fortran_:
2177 1.11 mrg _Interface_: 'subroutine acc_present_or_create(a)'
2178 1.11 mrg 'type, dimension(:[,:]...) :: a'
2179 1.11 mrg _Interface_: 'subroutine acc_present_or_create(a, len)'
2180 1.11 mrg 'type, dimension(:[,:]...) :: a'
2181 1.11 mrg 'integer len'
2182 1.11 mrg _Interface_: 'subroutine acc_pcreate(a)'
2183 1.11 mrg 'type, dimension(:[,:]...) :: a'
2184 1.11 mrg _Interface_: 'subroutine acc_pcreate(a, len)'
2185 1.11 mrg 'type, dimension(:[,:]...) :: a'
2186 1.11 mrg 'integer len'
2187 1.7 mrg
2188 1.7 mrg _Reference_:
2189 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2190 1.7 mrg 3.2.20.
2191 1.7 mrg
2192 1.7 mrg
2193 1.7 mrg File: libgomp.info, Node: acc_copyout, Next: acc_delete, Prev: acc_present_or_create, Up: OpenACC Runtime Library Routines
2194 1.7 mrg
2195 1.11 mrg 5.21 'acc_copyout' - Copy device memory to host memory.
2196 1.7 mrg =======================================================
2197 1.7 mrg
2198 1.7 mrg _Description_
2199 1.7 mrg This function copies mapped device memory to host memory which is
2200 1.7 mrg specified by host address A for a length LEN bytes in C/C++.
2201 1.7 mrg
2202 1.11 mrg In Fortran, two (2) forms are supported. In the first form, A
2203 1.11 mrg specifies a contiguous array section. The second form A specifies
2204 1.7 mrg a variable or array element and LEN specifies the length in bytes.
2205 1.7 mrg
2206 1.7 mrg _C/C++_:
2207 1.11 mrg _Prototype_: 'acc_copyout(h_void *a, size_t len);'
2208 1.7 mrg
2209 1.7 mrg _Fortran_:
2210 1.11 mrg _Interface_: 'subroutine acc_copyout(a)'
2211 1.11 mrg 'type, dimension(:[,:]...) :: a'
2212 1.11 mrg _Interface_: 'subroutine acc_copyout(a, len)'
2213 1.11 mrg 'type, dimension(:[,:]...) :: a'
2214 1.11 mrg 'integer len'
2215 1.7 mrg
2216 1.7 mrg _Reference_:
2217 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2218 1.7 mrg 3.2.21.
2219 1.7 mrg
2220 1.7 mrg
2221 1.7 mrg File: libgomp.info, Node: acc_delete, Next: acc_update_device, Prev: acc_copyout, Up: OpenACC Runtime Library Routines
2222 1.7 mrg
2223 1.11 mrg 5.22 'acc_delete' - Free device memory.
2224 1.7 mrg =======================================
2225 1.7 mrg
2226 1.7 mrg _Description_
2227 1.7 mrg This function frees previously allocated device memory specified by
2228 1.7 mrg the device address A and the length of LEN bytes.
2229 1.7 mrg
2230 1.11 mrg In Fortran, two (2) forms are supported. In the first form, A
2231 1.11 mrg specifies a contiguous array section. The second form A specifies
2232 1.7 mrg a variable or array element and LEN specifies the length in bytes.
2233 1.7 mrg
2234 1.7 mrg _C/C++_:
2235 1.11 mrg _Prototype_: 'acc_delete(h_void *a, size_t len);'
2236 1.7 mrg
2237 1.7 mrg _Fortran_:
2238 1.11 mrg _Interface_: 'subroutine acc_delete(a)'
2239 1.11 mrg 'type, dimension(:[,:]...) :: a'
2240 1.11 mrg _Interface_: 'subroutine acc_delete(a, len)'
2241 1.11 mrg 'type, dimension(:[,:]...) :: a'
2242 1.11 mrg 'integer len'
2243 1.7 mrg
2244 1.7 mrg _Reference_:
2245 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2246 1.7 mrg 3.2.22.
2247 1.7 mrg
2248 1.7 mrg
2249 1.7 mrg File: libgomp.info, Node: acc_update_device, Next: acc_update_self, Prev: acc_delete, Up: OpenACC Runtime Library Routines
2250 1.7 mrg
2251 1.11 mrg 5.23 'acc_update_device' - Update device memory from mapped host memory.
2252 1.7 mrg ========================================================================
2253 1.7 mrg
2254 1.7 mrg _Description_
2255 1.7 mrg This function updates the device copy from the previously mapped
2256 1.7 mrg host memory. The host memory is specified with the host address A
2257 1.7 mrg and a length of LEN bytes.
2258 1.7 mrg
2259 1.11 mrg In Fortran, two (2) forms are supported. In the first form, A
2260 1.11 mrg specifies a contiguous array section. The second form A specifies
2261 1.7 mrg a variable or array element and LEN specifies the length in bytes.
2262 1.7 mrg
2263 1.7 mrg _C/C++_:
2264 1.11 mrg _Prototype_: 'acc_update_device(h_void *a, size_t len);'
2265 1.7 mrg
2266 1.7 mrg _Fortran_:
2267 1.11 mrg _Interface_: 'subroutine acc_update_device(a)'
2268 1.11 mrg 'type, dimension(:[,:]...) :: a'
2269 1.11 mrg _Interface_: 'subroutine acc_update_device(a, len)'
2270 1.11 mrg 'type, dimension(:[,:]...) :: a'
2271 1.11 mrg 'integer len'
2272 1.7 mrg
2273 1.7 mrg _Reference_:
2274 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2275 1.7 mrg 3.2.23.
2276 1.7 mrg
2277 1.7 mrg
2278 1.7 mrg File: libgomp.info, Node: acc_update_self, Next: acc_map_data, Prev: acc_update_device, Up: OpenACC Runtime Library Routines
2279 1.7 mrg
2280 1.11 mrg 5.24 'acc_update_self' - Update host memory from mapped device memory.
2281 1.7 mrg ======================================================================
2282 1.7 mrg
2283 1.7 mrg _Description_
2284 1.7 mrg This function updates the host copy from the previously mapped
2285 1.7 mrg device memory. The host memory is specified with the host address
2286 1.7 mrg A and a length of LEN bytes.
2287 1.7 mrg
2288 1.11 mrg In Fortran, two (2) forms are supported. In the first form, A
2289 1.11 mrg specifies a contiguous array section. The second form A specifies
2290 1.7 mrg a variable or array element and LEN specifies the length in bytes.
2291 1.7 mrg
2292 1.7 mrg _C/C++_:
2293 1.11 mrg _Prototype_: 'acc_update_self(h_void *a, size_t len);'
2294 1.7 mrg
2295 1.7 mrg _Fortran_:
2296 1.11 mrg _Interface_: 'subroutine acc_update_self(a)'
2297 1.11 mrg 'type, dimension(:[,:]...) :: a'
2298 1.11 mrg _Interface_: 'subroutine acc_update_self(a, len)'
2299 1.11 mrg 'type, dimension(:[,:]...) :: a'
2300 1.11 mrg 'integer len'
2301 1.7 mrg
2302 1.7 mrg _Reference_:
2303 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2304 1.7 mrg 3.2.24.
2305 1.7 mrg
2306 1.7 mrg
2307 1.7 mrg File: libgomp.info, Node: acc_map_data, Next: acc_unmap_data, Prev: acc_update_self, Up: OpenACC Runtime Library Routines
2308 1.7 mrg
2309 1.11 mrg 5.25 'acc_map_data' - Map previously allocated device memory to host memory.
2310 1.7 mrg ============================================================================
2311 1.7 mrg
2312 1.7 mrg _Description_
2313 1.7 mrg This function maps previously allocated device and host memory.
2314 1.11 mrg The device memory is specified with the device address D. The host
2315 1.7 mrg memory is specified with the host address H and a length of LEN.
2316 1.7 mrg
2317 1.7 mrg _C/C++_:
2318 1.11 mrg _Prototype_: 'acc_map_data(h_void *h, d_void *d, size_t len);'
2319 1.7 mrg
2320 1.7 mrg _Reference_:
2321 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2322 1.7 mrg 3.2.25.
2323 1.7 mrg
2324 1.7 mrg
2325 1.7 mrg File: libgomp.info, Node: acc_unmap_data, Next: acc_deviceptr, Prev: acc_map_data, Up: OpenACC Runtime Library Routines
2326 1.7 mrg
2327 1.11 mrg 5.26 'acc_unmap_data' - Unmap device memory from host memory.
2328 1.7 mrg =============================================================
2329 1.7 mrg
2330 1.7 mrg _Description_
2331 1.11 mrg This function unmaps previously mapped device and host memory. The
2332 1.7 mrg latter specified by H.
2333 1.7 mrg
2334 1.7 mrg _C/C++_:
2335 1.11 mrg _Prototype_: 'acc_unmap_data(h_void *h);'
2336 1.7 mrg
2337 1.7 mrg _Reference_:
2338 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2339 1.7 mrg 3.2.26.
2340 1.7 mrg
2341 1.7 mrg
2342 1.7 mrg File: libgomp.info, Node: acc_deviceptr, Next: acc_hostptr, Prev: acc_unmap_data, Up: OpenACC Runtime Library Routines
2343 1.7 mrg
2344 1.11 mrg 5.27 'acc_deviceptr' - Get device pointer associated with specific host address.
2345 1.7 mrg ================================================================================
2346 1.7 mrg
2347 1.7 mrg _Description_
2348 1.7 mrg This function returns the device address that has been mapped to
2349 1.7 mrg the host address specified by H.
2350 1.7 mrg
2351 1.7 mrg _C/C++_:
2352 1.11 mrg _Prototype_: 'void *acc_deviceptr(h_void *h);'
2353 1.7 mrg
2354 1.7 mrg _Reference_:
2355 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2356 1.7 mrg 3.2.27.
2357 1.7 mrg
2358 1.7 mrg
2359 1.7 mrg File: libgomp.info, Node: acc_hostptr, Next: acc_is_present, Prev: acc_deviceptr, Up: OpenACC Runtime Library Routines
2360 1.7 mrg
2361 1.11 mrg 5.28 'acc_hostptr' - Get host pointer associated with specific device address.
2362 1.7 mrg ==============================================================================
2363 1.7 mrg
2364 1.7 mrg _Description_
2365 1.7 mrg This function returns the host address that has been mapped to the
2366 1.7 mrg device address specified by D.
2367 1.7 mrg
2368 1.7 mrg _C/C++_:
2369 1.11 mrg _Prototype_: 'void *acc_hostptr(d_void *d);'
2370 1.7 mrg
2371 1.7 mrg _Reference_:
2372 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2373 1.7 mrg 3.2.28.
2374 1.7 mrg
2375 1.7 mrg
2376 1.7 mrg File: libgomp.info, Node: acc_is_present, Next: acc_memcpy_to_device, Prev: acc_hostptr, Up: OpenACC Runtime Library Routines
2377 1.7 mrg
2378 1.11 mrg 5.29 'acc_is_present' - Indicate whether host variable / array is present on device.
2379 1.7 mrg ====================================================================================
2380 1.7 mrg
2381 1.7 mrg _Description_
2382 1.11 mrg This function indicates whether the specified host address in A and
2383 1.11 mrg a length of LEN bytes is present on the device. In C/C++, a
2384 1.7 mrg non-zero value is returned to indicate the presence of the mapped
2385 1.11 mrg memory on the device. A zero is returned to indicate the memory is
2386 1.7 mrg not mapped on the device.
2387 1.7 mrg
2388 1.11 mrg In Fortran, two (2) forms are supported. In the first form, A
2389 1.11 mrg specifies a contiguous array section. The second form A specifies
2390 1.7 mrg a variable or array element and LEN specifies the length in bytes.
2391 1.11 mrg If the host memory is mapped to device memory, then a 'true' is
2392 1.11 mrg returned. Otherwise, a 'false' is return to indicate the mapped
2393 1.7 mrg memory is not present.
2394 1.7 mrg
2395 1.7 mrg _C/C++_:
2396 1.11 mrg _Prototype_: 'int acc_is_present(h_void *a, size_t len);'
2397 1.7 mrg
2398 1.7 mrg _Fortran_:
2399 1.11 mrg _Interface_: 'function acc_is_present(a)'
2400 1.11 mrg 'type, dimension(:[,:]...) :: a'
2401 1.11 mrg 'logical acc_is_present'
2402 1.11 mrg _Interface_: 'function acc_is_present(a, len)'
2403 1.11 mrg 'type, dimension(:[,:]...) :: a'
2404 1.11 mrg 'integer len'
2405 1.11 mrg 'logical acc_is_present'
2406 1.7 mrg
2407 1.7 mrg _Reference_:
2408 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2409 1.7 mrg 3.2.29.
2410 1.7 mrg
2411 1.7 mrg
2412 1.7 mrg File: libgomp.info, Node: acc_memcpy_to_device, Next: acc_memcpy_from_device, Prev: acc_is_present, Up: OpenACC Runtime Library Routines
2413 1.7 mrg
2414 1.11 mrg 5.30 'acc_memcpy_to_device' - Copy host memory to device memory.
2415 1.7 mrg ================================================================
2416 1.7 mrg
2417 1.7 mrg _Description_
2418 1.7 mrg This function copies host memory specified by host address of SRC
2419 1.7 mrg to device memory specified by the device address DEST for a length
2420 1.7 mrg of BYTES bytes.
2421 1.7 mrg
2422 1.7 mrg _C/C++_:
2423 1.11 mrg _Prototype_: 'acc_memcpy_to_device(d_void *dest, h_void *src, size_t
2424 1.11 mrg bytes);'
2425 1.7 mrg
2426 1.7 mrg _Reference_:
2427 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2428 1.7 mrg 3.2.30.
2429 1.7 mrg
2430 1.7 mrg
2431 1.7 mrg File: libgomp.info, Node: acc_memcpy_from_device, Next: acc_get_current_cuda_device, Prev: acc_memcpy_to_device, Up: OpenACC Runtime Library Routines
2432 1.7 mrg
2433 1.11 mrg 5.31 'acc_memcpy_from_device' - Copy device memory to host memory.
2434 1.7 mrg ==================================================================
2435 1.7 mrg
2436 1.7 mrg _Description_
2437 1.7 mrg This function copies host memory specified by host address of SRC
2438 1.7 mrg from device memory specified by the device address DEST for a
2439 1.7 mrg length of BYTES bytes.
2440 1.7 mrg
2441 1.7 mrg _C/C++_:
2442 1.11 mrg _Prototype_: 'acc_memcpy_from_device(d_void *dest, h_void *src,
2443 1.11 mrg size_t bytes);'
2444 1.7 mrg
2445 1.7 mrg _Reference_:
2446 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2447 1.7 mrg 3.2.31.
2448 1.7 mrg
2449 1.7 mrg
2450 1.7 mrg File: libgomp.info, Node: acc_get_current_cuda_device, Next: acc_get_current_cuda_context, Prev: acc_memcpy_from_device, Up: OpenACC Runtime Library Routines
2451 1.7 mrg
2452 1.11 mrg 5.32 'acc_get_current_cuda_device' - Get CUDA device handle.
2453 1.7 mrg ============================================================
2454 1.7 mrg
2455 1.7 mrg _Description_
2456 1.11 mrg This function returns the CUDA device handle. This handle is the
2457 1.7 mrg same as used by the CUDA Runtime or Driver API's.
2458 1.7 mrg
2459 1.7 mrg _C/C++_:
2460 1.11 mrg _Prototype_: 'void *acc_get_current_cuda_device(void);'
2461 1.7 mrg
2462 1.7 mrg _Reference_:
2463 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2464 1.7 mrg A.2.1.1.
2465 1.7 mrg
2466 1.7 mrg
2467 1.7 mrg File: libgomp.info, Node: acc_get_current_cuda_context, Next: acc_get_cuda_stream, Prev: acc_get_current_cuda_device, Up: OpenACC Runtime Library Routines
2468 1.7 mrg
2469 1.11 mrg 5.33 'acc_get_current_cuda_context' - Get CUDA context handle.
2470 1.7 mrg ==============================================================
2471 1.7 mrg
2472 1.7 mrg _Description_
2473 1.11 mrg This function returns the CUDA context handle. This handle is the
2474 1.7 mrg same as used by the CUDA Runtime or Driver API's.
2475 1.7 mrg
2476 1.7 mrg _C/C++_:
2477 1.11 mrg _Prototype_: 'acc_get_current_cuda_context(void);'
2478 1.7 mrg
2479 1.7 mrg _Reference_:
2480 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2481 1.7 mrg A.2.1.2.
2482 1.7 mrg
2483 1.7 mrg
2484 1.7 mrg File: libgomp.info, Node: acc_get_cuda_stream, Next: acc_set_cuda_stream, Prev: acc_get_current_cuda_context, Up: OpenACC Runtime Library Routines
2485 1.7 mrg
2486 1.11 mrg 5.34 'acc_get_cuda_stream' - Get CUDA stream handle.
2487 1.7 mrg ====================================================
2488 1.7 mrg
2489 1.7 mrg _Description_
2490 1.11 mrg This function returns the CUDA stream handle. This handle is the
2491 1.7 mrg same as used by the CUDA Runtime or Driver API's.
2492 1.7 mrg
2493 1.7 mrg _C/C++_:
2494 1.11 mrg _Prototype_: 'acc_get_cuda_stream(void);'
2495 1.7 mrg
2496 1.7 mrg _Reference_:
2497 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2498 1.7 mrg A.2.1.3.
2499 1.7 mrg
2500 1.7 mrg
2501 1.7 mrg File: libgomp.info, Node: acc_set_cuda_stream, Prev: acc_get_cuda_stream, Up: OpenACC Runtime Library Routines
2502 1.7 mrg
2503 1.11 mrg 5.35 'acc_set_cuda_stream' - Set CUDA stream handle.
2504 1.7 mrg ====================================================
2505 1.7 mrg
2506 1.7 mrg _Description_
2507 1.7 mrg This function associates the stream handle specified by STREAM with
2508 1.7 mrg the asynchronous value specified by ASYNC.
2509 1.7 mrg
2510 1.7 mrg _C/C++_:
2511 1.11 mrg _Prototype_: 'acc_set_cuda_stream(int async void *stream);'
2512 1.7 mrg
2513 1.7 mrg _Reference_:
2514 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section
2515 1.7 mrg A.2.1.4.
2516 1.7 mrg
2517 1.7 mrg
2518 1.7 mrg File: libgomp.info, Node: OpenACC Environment Variables, Next: CUDA Streams Usage, Prev: OpenACC Runtime Library Routines, Up: Top
2519 1.7 mrg
2520 1.7 mrg 6 OpenACC Environment Variables
2521 1.7 mrg *******************************
2522 1.7 mrg
2523 1.11 mrg The variables 'ACC_DEVICE_TYPE' and 'ACC_DEVICE_NUM' are defined by
2524 1.7 mrg section 4 of the OpenACC specification in version 2.0. The variable
2525 1.11 mrg 'GCC_ACC_NOTIFY' is used for diagnostic purposes.
2526 1.7 mrg
2527 1.7 mrg * Menu:
2528 1.7 mrg
2529 1.7 mrg * ACC_DEVICE_TYPE::
2530 1.7 mrg * ACC_DEVICE_NUM::
2531 1.7 mrg * GCC_ACC_NOTIFY::
2532 1.7 mrg
2533 1.7 mrg
2534 1.7 mrg File: libgomp.info, Node: ACC_DEVICE_TYPE, Next: ACC_DEVICE_NUM, Up: OpenACC Environment Variables
2535 1.7 mrg
2536 1.11 mrg 6.1 'ACC_DEVICE_TYPE'
2537 1.7 mrg =====================
2538 1.7 mrg
2539 1.7 mrg _Reference_:
2540 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section 4.1.
2541 1.7 mrg
2542 1.7 mrg
2543 1.7 mrg File: libgomp.info, Node: ACC_DEVICE_NUM, Next: GCC_ACC_NOTIFY, Prev: ACC_DEVICE_TYPE, Up: OpenACC Environment Variables
2544 1.7 mrg
2545 1.11 mrg 6.2 'ACC_DEVICE_NUM'
2546 1.7 mrg ====================
2547 1.7 mrg
2548 1.7 mrg _Reference_:
2549 1.11 mrg OpenACC specification v2.0 (https://www.openacc.org), section 4.2.
2550 1.7 mrg
2551 1.7 mrg
2552 1.7 mrg File: libgomp.info, Node: GCC_ACC_NOTIFY, Prev: ACC_DEVICE_NUM, Up: OpenACC Environment Variables
2553 1.7 mrg
2554 1.11 mrg 6.3 'GCC_ACC_NOTIFY'
2555 1.7 mrg ====================
2556 1.7 mrg
2557 1.7 mrg _Description_:
2558 1.7 mrg Print debug information pertaining to the accelerator.
2559 1.7 mrg
2560 1.7 mrg
2561 1.7 mrg File: libgomp.info, Node: CUDA Streams Usage, Next: OpenACC Library Interoperability, Prev: OpenACC Environment Variables, Up: Top
2562 1.7 mrg
2563 1.7 mrg 7 CUDA Streams Usage
2564 1.7 mrg ********************
2565 1.7 mrg
2566 1.11 mrg This applies to the 'nvptx' plugin only.
2567 1.7 mrg
2568 1.7 mrg The library provides elements that perform asynchronous movement of
2569 1.7 mrg data and asynchronous operation of computing constructs. This
2570 1.7 mrg asynchronous functionality is implemented by making use of CUDA
2571 1.7 mrg streams(1).
2572 1.7 mrg
2573 1.7 mrg The primary means by that the asychronous functionality is accessed
2574 1.7 mrg is through the use of those OpenACC directives which make use of the
2575 1.11 mrg 'async' and 'wait' clauses. When the 'async' clause is first used with
2576 1.11 mrg a directive, it creates a CUDA stream. If an 'async-argument' is used
2577 1.11 mrg with the 'async' clause, then the stream is associated with the
2578 1.11 mrg specified 'async-argument'.
2579 1.7 mrg
2580 1.7 mrg Following the creation of an association between a CUDA stream and
2581 1.11 mrg the 'async-argument' of an 'async' clause, both the 'wait' clause and
2582 1.11 mrg the 'wait' directive can be used. When either the clause or directive
2583 1.7 mrg is used after stream creation, it creates a rendezvous point whereby
2584 1.7 mrg execution waits until all operations associated with the
2585 1.11 mrg 'async-argument', that is, stream, have completed.
2586 1.7 mrg
2587 1.7 mrg Normally, the management of the streams that are created as a result
2588 1.11 mrg of using the 'async' clause, is done without any intervention by the
2589 1.11 mrg caller. This implies the association between the 'async-argument' and
2590 1.7 mrg the CUDA stream will be maintained for the lifetime of the program.
2591 1.7 mrg However, this association can be changed through the use of the library
2592 1.11 mrg function 'acc_set_cuda_stream'. When the function 'acc_set_cuda_stream'
2593 1.11 mrg is called, the CUDA stream that was originally associated with the
2594 1.11 mrg 'async' clause will be destroyed. Caution should be taken when changing
2595 1.11 mrg the association as subsequent references to the 'async-argument' refer
2596 1.11 mrg to a different CUDA stream.
2597 1.7 mrg
2598 1.7 mrg ---------- Footnotes ----------
2599 1.7 mrg
2600 1.7 mrg (1) See "Stream Management" in "CUDA Driver API", TRM-06703-001,
2601 1.7 mrg Version 5.5, for additional information
2602 1.7 mrg
2603 1.7 mrg
2604 1.7 mrg File: libgomp.info, Node: OpenACC Library Interoperability, Next: The libgomp ABI, Prev: CUDA Streams Usage, Up: Top
2605 1.7 mrg
2606 1.7 mrg 8 OpenACC Library Interoperability
2607 1.7 mrg **********************************
2608 1.7 mrg
2609 1.7 mrg 8.1 Introduction
2610 1.7 mrg ================
2611 1.7 mrg
2612 1.7 mrg The OpenACC library uses the CUDA Driver API, and may interact with
2613 1.11 mrg programs that use the Runtime library directly, or another library based
2614 1.11 mrg on the Runtime library, e.g., CUBLAS(1). This chapter describes the use
2615 1.11 mrg cases and what changes are required in order to use both the OpenACC
2616 1.11 mrg library and the CUBLAS and Runtime libraries within a program.
2617 1.7 mrg
2618 1.7 mrg 8.2 First invocation: NVIDIA CUBLAS library API
2619 1.7 mrg ===============================================
2620 1.7 mrg
2621 1.7 mrg In this first use case (see below), a function in the CUBLAS library is
2622 1.11 mrg called prior to any of the functions in the OpenACC library. More
2623 1.11 mrg specifically, the function 'cublasCreate()'.
2624 1.7 mrg
2625 1.7 mrg When invoked, the function initializes the library and allocates the
2626 1.7 mrg hardware resources on the host and the device on behalf of the caller.
2627 1.7 mrg Once the initialization and allocation has completed, a handle is
2628 1.11 mrg returned to the caller. The OpenACC library also requires
2629 1.11 mrg initialization and allocation of hardware resources. Since the CUBLAS
2630 1.11 mrg library has already allocated the hardware resources for the device, all
2631 1.11 mrg that is left to do is to initialize the OpenACC library and acquire the
2632 1.11 mrg hardware resources on the host.
2633 1.7 mrg
2634 1.7 mrg Prior to calling the OpenACC function that initializes the library
2635 1.11 mrg and allocate the host hardware resources, you need to acquire the device
2636 1.11 mrg number that was allocated during the call to 'cublasCreate()'. The
2637 1.11 mrg invoking of the runtime library function 'cudaGetDevice()' accomplishes
2638 1.11 mrg this. Once acquired, the device number is passed along with the device
2639 1.11 mrg type as parameters to the OpenACC library function
2640 1.11 mrg 'acc_set_device_num()'.
2641 1.11 mrg
2642 1.11 mrg Once the call to 'acc_set_device_num()' has completed, the OpenACC
2643 1.11 mrg library uses the context that was created during the call to
2644 1.11 mrg 'cublasCreate()'. In other words, both libraries will be sharing the
2645 1.7 mrg same context.
2646 1.7 mrg
2647 1.7 mrg /* Create the handle */
2648 1.7 mrg s = cublasCreate(&h);
2649 1.7 mrg if (s != CUBLAS_STATUS_SUCCESS)
2650 1.7 mrg {
2651 1.7 mrg fprintf(stderr, "cublasCreate failed %d\n", s);
2652 1.7 mrg exit(EXIT_FAILURE);
2653 1.7 mrg }
2654 1.7 mrg
2655 1.7 mrg /* Get the device number */
2656 1.7 mrg e = cudaGetDevice(&dev);
2657 1.7 mrg if (e != cudaSuccess)
2658 1.7 mrg {
2659 1.7 mrg fprintf(stderr, "cudaGetDevice failed %d\n", e);
2660 1.7 mrg exit(EXIT_FAILURE);
2661 1.7 mrg }
2662 1.7 mrg
2663 1.7 mrg /* Initialize OpenACC library and use device 'dev' */
2664 1.7 mrg acc_set_device_num(dev, acc_device_nvidia);
2665 1.11 mrg
2666 1.11 mrg Use Case 1
2667 1.7 mrg
2668 1.7 mrg 8.3 First invocation: OpenACC library API
2669 1.7 mrg =========================================
2670 1.7 mrg
2671 1.7 mrg In this second use case (see below), a function in the OpenACC library
2672 1.11 mrg is called prior to any of the functions in the CUBLAS library. More
2673 1.11 mrg specificially, the function 'acc_set_device_num()'.
2674 1.7 mrg
2675 1.11 mrg In the use case presented here, the function 'acc_set_device_num()'
2676 1.7 mrg is used to both initialize the OpenACC library and allocate the hardware
2677 1.11 mrg resources on the host and the device. In the call to the function, the
2678 1.11 mrg call parameters specify which device to use and what device type to use,
2679 1.11 mrg i.e., 'acc_device_nvidia'. It should be noted that this is but one
2680 1.7 mrg method to initialize the OpenACC library and allocate the appropriate
2681 1.11 mrg hardware resources. Other methods are available through the use of
2682 1.7 mrg environment variables and these will be discussed in the next section.
2683 1.7 mrg
2684 1.11 mrg Once the call to 'acc_set_device_num()' has completed, other OpenACC
2685 1.7 mrg functions can be called as seen with multiple calls being made to
2686 1.11 mrg 'acc_copyin()'. In addition, calls can be made to functions in the
2687 1.11 mrg CUBLAS library. In the use case a call to 'cublasCreate()' is made
2688 1.11 mrg subsequent to the calls to 'acc_copyin()'. As seen in the previous use
2689 1.11 mrg case, a call to 'cublasCreate()' initializes the CUBLAS library and
2690 1.7 mrg allocates the hardware resources on the host and the device. However,
2691 1.11 mrg since the device has already been allocated, 'cublasCreate()' will only
2692 1.7 mrg initialize the CUBLAS library and allocate the appropriate hardware
2693 1.11 mrg resources on the host. The context that was created as part of the
2694 1.7 mrg OpenACC initialization is shared with the CUBLAS library, similarly to
2695 1.7 mrg the first use case.
2696 1.7 mrg
2697 1.7 mrg dev = 0;
2698 1.7 mrg
2699 1.7 mrg acc_set_device_num(dev, acc_device_nvidia);
2700 1.7 mrg
2701 1.7 mrg /* Copy the first set to the device */
2702 1.7 mrg d_X = acc_copyin(&h_X[0], N * sizeof (float));
2703 1.7 mrg if (d_X == NULL)
2704 1.7 mrg {
2705 1.7 mrg fprintf(stderr, "copyin error h_X\n");
2706 1.7 mrg exit(EXIT_FAILURE);
2707 1.7 mrg }
2708 1.7 mrg
2709 1.7 mrg /* Copy the second set to the device */
2710 1.7 mrg d_Y = acc_copyin(&h_Y1[0], N * sizeof (float));
2711 1.7 mrg if (d_Y == NULL)
2712 1.7 mrg {
2713 1.7 mrg fprintf(stderr, "copyin error h_Y1\n");
2714 1.7 mrg exit(EXIT_FAILURE);
2715 1.7 mrg }
2716 1.7 mrg
2717 1.7 mrg /* Create the handle */
2718 1.7 mrg s = cublasCreate(&h);
2719 1.7 mrg if (s != CUBLAS_STATUS_SUCCESS)
2720 1.7 mrg {
2721 1.7 mrg fprintf(stderr, "cublasCreate failed %d\n", s);
2722 1.7 mrg exit(EXIT_FAILURE);
2723 1.7 mrg }
2724 1.7 mrg
2725 1.7 mrg /* Perform saxpy using CUBLAS library function */
2726 1.7 mrg s = cublasSaxpy(h, N, &alpha, d_X, 1, d_Y, 1);
2727 1.7 mrg if (s != CUBLAS_STATUS_SUCCESS)
2728 1.7 mrg {
2729 1.7 mrg fprintf(stderr, "cublasSaxpy failed %d\n", s);
2730 1.7 mrg exit(EXIT_FAILURE);
2731 1.7 mrg }
2732 1.7 mrg
2733 1.7 mrg /* Copy the results from the device */
2734 1.7 mrg acc_memcpy_from_device(&h_Y1[0], d_Y, N * sizeof (float));
2735 1.11 mrg
2736 1.11 mrg Use Case 2
2737 1.7 mrg
2738 1.7 mrg 8.4 OpenACC library and environment variables
2739 1.7 mrg =============================================
2740 1.7 mrg
2741 1.7 mrg There are two environment variables associated with the OpenACC library
2742 1.7 mrg that may be used to control the device type and device number:
2743 1.11 mrg 'ACC_DEVICE_TYPE' and 'ACC_DEVICE_NUM', respecively. These two
2744 1.7 mrg environement variables can be used as an alternative to calling
2745 1.11 mrg 'acc_set_device_num()'. As seen in the second use case, the device type
2746 1.11 mrg and device number were specified using 'acc_set_device_num()'. If
2747 1.7 mrg however, the aforementioned environment variables were set, then the
2748 1.11 mrg call to 'acc_set_device_num()' would not be required.
2749 1.7 mrg
2750 1.11 mrg The use of the environment variables is only relevant when an OpenACC
2751 1.11 mrg function is called prior to a call to 'cudaCreate()'. If 'cudaCreate()'
2752 1.11 mrg is called prior to a call to an OpenACC function, then you must call
2753 1.11 mrg 'acc_set_device_num()'(2)
2754 1.7 mrg
2755 1.7 mrg ---------- Footnotes ----------
2756 1.7 mrg
2757 1.7 mrg (1) See section 2.26, "Interactions with the CUDA Driver API" in
2758 1.7 mrg "CUDA Runtime API", Version 5.5, and section 2.27, "VDPAU
2759 1.11 mrg Interoperability", in "CUDA Driver API", TRM-06703-001, Version 5.5, for
2760 1.11 mrg additional information on library interoperability.
2761 1.7 mrg
2762 1.11 mrg (2) More complete information about 'ACC_DEVICE_TYPE' and
2763 1.11 mrg 'ACC_DEVICE_NUM' can be found in sections 4.1 and 4.2 of the OpenACC
2764 1.11 mrg (https://www.openacc.org) Application Programming Interface, Version
2765 1.7 mrg 2.0.
2766 1.7 mrg
2767 1.7 mrg
2768 1.7 mrg File: libgomp.info, Node: The libgomp ABI, Next: Reporting Bugs, Prev: OpenACC Library Interoperability, Up: Top
2769 1.7 mrg
2770 1.7 mrg 9 The libgomp ABI
2771 1.1 mrg *****************
2772 1.1 mrg
2773 1.11 mrg The following sections present notes on the external ABI as presented by
2774 1.11 mrg libgomp. Only maintainers should need them.
2775 1.1 mrg
2776 1.1 mrg * Menu:
2777 1.1 mrg
2778 1.1 mrg * Implementing MASTER construct::
2779 1.1 mrg * Implementing CRITICAL construct::
2780 1.1 mrg * Implementing ATOMIC construct::
2781 1.1 mrg * Implementing FLUSH construct::
2782 1.1 mrg * Implementing BARRIER construct::
2783 1.1 mrg * Implementing THREADPRIVATE construct::
2784 1.1 mrg * Implementing PRIVATE clause::
2785 1.1 mrg * Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses::
2786 1.1 mrg * Implementing REDUCTION clause::
2787 1.1 mrg * Implementing PARALLEL construct::
2788 1.1 mrg * Implementing FOR construct::
2789 1.1 mrg * Implementing ORDERED construct::
2790 1.1 mrg * Implementing SECTIONS construct::
2791 1.1 mrg * Implementing SINGLE construct::
2792 1.7 mrg * Implementing OpenACC's PARALLEL construct::
2793 1.1 mrg
2794 1.1 mrg
2795 1.1 mrg File: libgomp.info, Node: Implementing MASTER construct, Next: Implementing CRITICAL construct, Up: The libgomp ABI
2796 1.1 mrg
2797 1.7 mrg 9.1 Implementing MASTER construct
2798 1.1 mrg =================================
2799 1.1 mrg
2800 1.1 mrg if (omp_get_thread_num () == 0)
2801 1.1 mrg block
2802 1.1 mrg
2803 1.1 mrg Alternately, we generate two copies of the parallel subfunction and
2804 1.3 mrg only include this in the version run by the master thread. Surely this
2805 1.3 mrg is not worthwhile though...
2806 1.1 mrg
2807 1.1 mrg
2808 1.1 mrg File: libgomp.info, Node: Implementing CRITICAL construct, Next: Implementing ATOMIC construct, Prev: Implementing MASTER construct, Up: The libgomp ABI
2809 1.1 mrg
2810 1.7 mrg 9.2 Implementing CRITICAL construct
2811 1.1 mrg ===================================
2812 1.1 mrg
2813 1.1 mrg Without a specified name,
2814 1.1 mrg
2815 1.1 mrg void GOMP_critical_start (void);
2816 1.1 mrg void GOMP_critical_end (void);
2817 1.1 mrg
2818 1.1 mrg so that we don't get COPY relocations from libgomp to the main
2819 1.1 mrg application.
2820 1.1 mrg
2821 1.1 mrg With a specified name, use omp_set_lock and omp_unset_lock with name
2822 1.1 mrg being transformed into a variable declared like
2823 1.1 mrg
2824 1.1 mrg omp_lock_t gomp_critical_user_<name> __attribute__((common))
2825 1.1 mrg
2826 1.1 mrg Ideally the ABI would specify that all zero is a valid unlocked
2827 1.3 mrg state, and so we wouldn't need to initialize this at startup.
2828 1.1 mrg
2829 1.1 mrg
2830 1.1 mrg File: libgomp.info, Node: Implementing ATOMIC construct, Next: Implementing FLUSH construct, Prev: Implementing CRITICAL construct, Up: The libgomp ABI
2831 1.1 mrg
2832 1.7 mrg 9.3 Implementing ATOMIC construct
2833 1.1 mrg =================================
2834 1.1 mrg
2835 1.11 mrg The target should implement the '__sync' builtins.
2836 1.1 mrg
2837 1.1 mrg Failing that we could add
2838 1.1 mrg
2839 1.1 mrg void GOMP_atomic_enter (void)
2840 1.1 mrg void GOMP_atomic_exit (void)
2841 1.1 mrg
2842 1.1 mrg which reuses the regular lock code, but with yet another lock object
2843 1.1 mrg private to the library.
2844 1.1 mrg
2845 1.1 mrg
2846 1.1 mrg File: libgomp.info, Node: Implementing FLUSH construct, Next: Implementing BARRIER construct, Prev: Implementing ATOMIC construct, Up: The libgomp ABI
2847 1.1 mrg
2848 1.7 mrg 9.4 Implementing FLUSH construct
2849 1.1 mrg ================================
2850 1.1 mrg
2851 1.11 mrg Expands to the '__sync_synchronize' builtin.
2852 1.1 mrg
2853 1.1 mrg
2854 1.1 mrg File: libgomp.info, Node: Implementing BARRIER construct, Next: Implementing THREADPRIVATE construct, Prev: Implementing FLUSH construct, Up: The libgomp ABI
2855 1.1 mrg
2856 1.7 mrg 9.5 Implementing BARRIER construct
2857 1.1 mrg ==================================
2858 1.1 mrg
2859 1.1 mrg void GOMP_barrier (void)
2860 1.1 mrg
2861 1.1 mrg
2862 1.1 mrg File: libgomp.info, Node: Implementing THREADPRIVATE construct, Next: Implementing PRIVATE clause, Prev: Implementing BARRIER construct, Up: The libgomp ABI
2863 1.1 mrg
2864 1.7 mrg 9.6 Implementing THREADPRIVATE construct
2865 1.1 mrg ========================================
2866 1.1 mrg
2867 1.11 mrg In _most_ cases we can map this directly to '__thread'. Except that OMP
2868 1.11 mrg allows constructors for C++ objects. We can either refuse to support
2869 1.11 mrg this (how often is it used?) or we can implement something akin to
2870 1.11 mrg .ctors.
2871 1.1 mrg
2872 1.1 mrg Even more ideally, this ctor feature is handled by extensions to the
2873 1.1 mrg main pthreads library. Failing that, we can have a set of entry points
2874 1.1 mrg to register ctor functions to be called.
2875 1.1 mrg
2876 1.1 mrg
2877 1.1 mrg File: libgomp.info, Node: Implementing PRIVATE clause, Next: Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses, Prev: Implementing THREADPRIVATE construct, Up: The libgomp ABI
2878 1.1 mrg
2879 1.7 mrg 9.7 Implementing PRIVATE clause
2880 1.1 mrg ===============================
2881 1.1 mrg
2882 1.1 mrg In association with a PARALLEL, or within the lexical extent of a
2883 1.1 mrg PARALLEL block, the variable becomes a local variable in the parallel
2884 1.1 mrg subfunction.
2885 1.1 mrg
2886 1.1 mrg In association with FOR or SECTIONS blocks, create a new automatic
2887 1.1 mrg variable within the current function. This preserves the semantic of
2888 1.1 mrg new variable creation.
2889 1.1 mrg
2890 1.1 mrg
2891 1.1 mrg File: libgomp.info, Node: Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses, Next: Implementing REDUCTION clause, Prev: Implementing PRIVATE clause, Up: The libgomp ABI
2892 1.1 mrg
2893 1.7 mrg 9.8 Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses
2894 1.1 mrg ========================================================================
2895 1.1 mrg
2896 1.3 mrg This seems simple enough for PARALLEL blocks. Create a private struct
2897 1.3 mrg for communicating between the parent and subfunction. In the parent,
2898 1.3 mrg copy in values for scalar and "small" structs; copy in addresses for
2899 1.3 mrg others TREE_ADDRESSABLE types. In the subfunction, copy the value into
2900 1.3 mrg the local variable.
2901 1.1 mrg
2902 1.11 mrg It is not clear what to do with bare FOR or SECTION blocks. The only
2903 1.11 mrg thing I can figure is that we do something like:
2904 1.1 mrg
2905 1.1 mrg #pragma omp for firstprivate(x) lastprivate(y)
2906 1.1 mrg for (int i = 0; i < n; ++i)
2907 1.1 mrg body;
2908 1.1 mrg
2909 1.1 mrg which becomes
2910 1.1 mrg
2911 1.1 mrg {
2912 1.1 mrg int x = x, y;
2913 1.1 mrg
2914 1.1 mrg // for stuff
2915 1.1 mrg
2916 1.1 mrg if (i == n)
2917 1.1 mrg y = y;
2918 1.1 mrg }
2919 1.1 mrg
2920 1.1 mrg where the "x=x" and "y=y" assignments actually have different uids
2921 1.11 mrg for the two variables, i.e. not something you could write directly in
2922 1.11 mrg C. Presumably this only makes sense if the "outer" x and y are global
2923 1.1 mrg variables.
2924 1.1 mrg
2925 1.1 mrg COPYPRIVATE would work the same way, except the structure broadcast
2926 1.1 mrg would have to happen via SINGLE machinery instead.
2927 1.1 mrg
2928 1.1 mrg
2929 1.1 mrg File: libgomp.info, Node: Implementing REDUCTION clause, Next: Implementing PARALLEL construct, Prev: Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses, Up: The libgomp ABI
2930 1.1 mrg
2931 1.7 mrg 9.9 Implementing REDUCTION clause
2932 1.1 mrg =================================
2933 1.1 mrg
2934 1.1 mrg The private struct mentioned in the previous section should have a
2935 1.11 mrg pointer to an array of the type of the variable, indexed by the thread's
2936 1.11 mrg TEAM_ID. The thread stores its final value into the array, and after
2937 1.11 mrg the barrier, the master thread iterates over the array to collect the
2938 1.11 mrg values.
2939 1.1 mrg
2940 1.1 mrg
2941 1.1 mrg File: libgomp.info, Node: Implementing PARALLEL construct, Next: Implementing FOR construct, Prev: Implementing REDUCTION clause, Up: The libgomp ABI
2942 1.1 mrg
2943 1.7 mrg 9.10 Implementing PARALLEL construct
2944 1.1 mrg ====================================
2945 1.1 mrg
2946 1.1 mrg #pragma omp parallel
2947 1.1 mrg {
2948 1.1 mrg body;
2949 1.1 mrg }
2950 1.1 mrg
2951 1.1 mrg becomes
2952 1.1 mrg
2953 1.1 mrg void subfunction (void *data)
2954 1.1 mrg {
2955 1.1 mrg use data;
2956 1.1 mrg body;
2957 1.1 mrg }
2958 1.1 mrg
2959 1.1 mrg setup data;
2960 1.1 mrg GOMP_parallel_start (subfunction, &data, num_threads);
2961 1.1 mrg subfunction (&data);
2962 1.1 mrg GOMP_parallel_end ();
2963 1.1 mrg
2964 1.1 mrg void GOMP_parallel_start (void (*fn)(void *), void *data, unsigned num_threads)
2965 1.1 mrg
2966 1.1 mrg The FN argument is the subfunction to be run in parallel.
2967 1.1 mrg
2968 1.1 mrg The DATA argument is a pointer to a structure used to communicate
2969 1.1 mrg data in and out of the subfunction, as discussed above with respect to
2970 1.1 mrg FIRSTPRIVATE et al.
2971 1.1 mrg
2972 1.1 mrg The NUM_THREADS argument is 1 if an IF clause is present and false,
2973 1.1 mrg or the value of the NUM_THREADS clause, if present, or 0.
2974 1.1 mrg
2975 1.11 mrg The function needs to create the appropriate number of threads and/or
2976 1.11 mrg launch them from the dock. It needs to create the team structure and
2977 1.11 mrg assign team ids.
2978 1.1 mrg
2979 1.1 mrg void GOMP_parallel_end (void)
2980 1.1 mrg
2981 1.1 mrg Tears down the team and returns us to the previous
2982 1.11 mrg 'omp_in_parallel()' state.
2983 1.1 mrg
2984 1.1 mrg
2985 1.1 mrg File: libgomp.info, Node: Implementing FOR construct, Next: Implementing ORDERED construct, Prev: Implementing PARALLEL construct, Up: The libgomp ABI
2986 1.1 mrg
2987 1.7 mrg 9.11 Implementing FOR construct
2988 1.1 mrg ===============================
2989 1.1 mrg
2990 1.1 mrg #pragma omp parallel for
2991 1.1 mrg for (i = lb; i <= ub; i++)
2992 1.1 mrg body;
2993 1.1 mrg
2994 1.1 mrg becomes
2995 1.1 mrg
2996 1.1 mrg void subfunction (void *data)
2997 1.1 mrg {
2998 1.1 mrg long _s0, _e0;
2999 1.1 mrg while (GOMP_loop_static_next (&_s0, &_e0))
3000 1.1 mrg {
3001 1.1 mrg long _e1 = _e0, i;
3002 1.1 mrg for (i = _s0; i < _e1; i++)
3003 1.1 mrg body;
3004 1.1 mrg }
3005 1.1 mrg GOMP_loop_end_nowait ();
3006 1.1 mrg }
3007 1.1 mrg
3008 1.1 mrg GOMP_parallel_loop_static (subfunction, NULL, 0, lb, ub+1, 1, 0);
3009 1.1 mrg subfunction (NULL);
3010 1.1 mrg GOMP_parallel_end ();
3011 1.1 mrg
3012 1.1 mrg #pragma omp for schedule(runtime)
3013 1.1 mrg for (i = 0; i < n; i++)
3014 1.1 mrg body;
3015 1.1 mrg
3016 1.1 mrg becomes
3017 1.1 mrg
3018 1.1 mrg {
3019 1.1 mrg long i, _s0, _e0;
3020 1.1 mrg if (GOMP_loop_runtime_start (0, n, 1, &_s0, &_e0))
3021 1.1 mrg do {
3022 1.1 mrg long _e1 = _e0;
3023 1.1 mrg for (i = _s0, i < _e0; i++)
3024 1.1 mrg body;
3025 1.1 mrg } while (GOMP_loop_runtime_next (&_s0, _&e0));
3026 1.1 mrg GOMP_loop_end ();
3027 1.1 mrg }
3028 1.1 mrg
3029 1.3 mrg Note that while it looks like there is trickiness to propagating a
3030 1.1 mrg non-constant STEP, there isn't really. We're explicitly allowed to
3031 1.1 mrg evaluate it as many times as we want, and any variables involved should
3032 1.1 mrg automatically be handled as PRIVATE or SHARED like any other variables.
3033 1.1 mrg So the expression should remain evaluable in the subfunction. We can
3034 1.11 mrg also pull it into a local variable if we like, but since its supposed to
3035 1.11 mrg remain unchanged, we can also not if we like.
3036 1.1 mrg
3037 1.11 mrg If we have SCHEDULE(STATIC), and no ORDERED, then we ought to be able
3038 1.11 mrg to get away with no work-sharing context at all, since we can simply
3039 1.11 mrg perform the arithmetic directly in each thread to divide up the
3040 1.11 mrg iterations. Which would mean that we wouldn't need to call any of these
3041 1.11 mrg routines.
3042 1.1 mrg
3043 1.1 mrg There are separate routines for handling loops with an ORDERED
3044 1.1 mrg clause. Bookkeeping for that is non-trivial...
3045 1.1 mrg
3046 1.1 mrg
3047 1.1 mrg File: libgomp.info, Node: Implementing ORDERED construct, Next: Implementing SECTIONS construct, Prev: Implementing FOR construct, Up: The libgomp ABI
3048 1.1 mrg
3049 1.7 mrg 9.12 Implementing ORDERED construct
3050 1.1 mrg ===================================
3051 1.1 mrg
3052 1.1 mrg void GOMP_ordered_start (void)
3053 1.1 mrg void GOMP_ordered_end (void)
3054 1.1 mrg
3055 1.1 mrg
3056 1.1 mrg File: libgomp.info, Node: Implementing SECTIONS construct, Next: Implementing SINGLE construct, Prev: Implementing ORDERED construct, Up: The libgomp ABI
3057 1.1 mrg
3058 1.7 mrg 9.13 Implementing SECTIONS construct
3059 1.1 mrg ====================================
3060 1.1 mrg
3061 1.1 mrg A block as
3062 1.1 mrg
3063 1.1 mrg #pragma omp sections
3064 1.1 mrg {
3065 1.1 mrg #pragma omp section
3066 1.1 mrg stmt1;
3067 1.1 mrg #pragma omp section
3068 1.1 mrg stmt2;
3069 1.1 mrg #pragma omp section
3070 1.1 mrg stmt3;
3071 1.1 mrg }
3072 1.1 mrg
3073 1.1 mrg becomes
3074 1.1 mrg
3075 1.1 mrg for (i = GOMP_sections_start (3); i != 0; i = GOMP_sections_next ())
3076 1.1 mrg switch (i)
3077 1.1 mrg {
3078 1.1 mrg case 1:
3079 1.1 mrg stmt1;
3080 1.1 mrg break;
3081 1.1 mrg case 2:
3082 1.1 mrg stmt2;
3083 1.1 mrg break;
3084 1.1 mrg case 3:
3085 1.1 mrg stmt3;
3086 1.1 mrg break;
3087 1.1 mrg }
3088 1.1 mrg GOMP_barrier ();
3089 1.1 mrg
3090 1.1 mrg
3091 1.7 mrg File: libgomp.info, Node: Implementing SINGLE construct, Next: Implementing OpenACC's PARALLEL construct, Prev: Implementing SECTIONS construct, Up: The libgomp ABI
3092 1.1 mrg
3093 1.7 mrg 9.14 Implementing SINGLE construct
3094 1.1 mrg ==================================
3095 1.1 mrg
3096 1.1 mrg A block like
3097 1.1 mrg
3098 1.1 mrg #pragma omp single
3099 1.1 mrg {
3100 1.1 mrg body;
3101 1.1 mrg }
3102 1.1 mrg
3103 1.1 mrg becomes
3104 1.1 mrg
3105 1.1 mrg if (GOMP_single_start ())
3106 1.1 mrg body;
3107 1.1 mrg GOMP_barrier ();
3108 1.1 mrg
3109 1.1 mrg while
3110 1.1 mrg
3111 1.1 mrg #pragma omp single copyprivate(x)
3112 1.1 mrg body;
3113 1.1 mrg
3114 1.1 mrg becomes
3115 1.1 mrg
3116 1.1 mrg datap = GOMP_single_copy_start ();
3117 1.1 mrg if (datap == NULL)
3118 1.1 mrg {
3119 1.1 mrg body;
3120 1.1 mrg data.x = x;
3121 1.1 mrg GOMP_single_copy_end (&data);
3122 1.1 mrg }
3123 1.1 mrg else
3124 1.1 mrg x = datap->x;
3125 1.1 mrg GOMP_barrier ();
3126 1.1 mrg
3127 1.1 mrg
3128 1.7 mrg File: libgomp.info, Node: Implementing OpenACC's PARALLEL construct, Prev: Implementing SINGLE construct, Up: The libgomp ABI
3129 1.7 mrg
3130 1.7 mrg 9.15 Implementing OpenACC's PARALLEL construct
3131 1.7 mrg ==============================================
3132 1.7 mrg
3133 1.7 mrg void GOACC_parallel ()
3134 1.7 mrg
3135 1.7 mrg
3136 1.1 mrg File: libgomp.info, Node: Reporting Bugs, Next: Copying, Prev: The libgomp ABI, Up: Top
3137 1.1 mrg
3138 1.7 mrg 10 Reporting Bugs
3139 1.7 mrg *****************
3140 1.1 mrg
3141 1.5 mrg Bugs in the GNU Offloading and Multi Processing Runtime Library should
3142 1.5 mrg be reported via Bugzilla (http://gcc.gnu.org/bugzilla/). Please add
3143 1.11 mrg "openacc", or "openmp", or both to the keywords field in the bug report,
3144 1.11 mrg as appropriate.
3145 1.1 mrg
3146 1.1 mrg
3147 1.1 mrg File: libgomp.info, Node: Copying, Next: GNU Free Documentation License, Prev: Reporting Bugs, Up: Top
3148 1.1 mrg
3149 1.3 mrg GNU General Public License
3150 1.1 mrg **************************
3151 1.1 mrg
3152 1.3 mrg Version 3, 29 June 2007
3153 1.1 mrg
3154 1.11 mrg Copyright (C) 2007 Free Software Foundation, Inc. <http://fsf.org/>
3155 1.1 mrg
3156 1.3 mrg Everyone is permitted to copy and distribute verbatim copies of this
3157 1.3 mrg license document, but changing it is not allowed.
3158 1.1 mrg
3159 1.1 mrg Preamble
3160 1.1 mrg ========
3161 1.1 mrg
3162 1.3 mrg The GNU General Public License is a free, copyleft license for software
3163 1.3 mrg and other kinds of works.
3164 1.3 mrg
3165 1.3 mrg The licenses for most software and other practical works are designed
3166 1.3 mrg to take away your freedom to share and change the works. By contrast,
3167 1.3 mrg the GNU General Public License is intended to guarantee your freedom to
3168 1.3 mrg share and change all versions of a program-to make sure it remains free
3169 1.3 mrg software for all its users. We, the Free Software Foundation, use the
3170 1.3 mrg GNU General Public License for most of our software; it applies also to
3171 1.3 mrg any other work released this way by its authors. You can apply it to
3172 1.1 mrg your programs, too.
3173 1.1 mrg
3174 1.1 mrg When we speak of free software, we are referring to freedom, not
3175 1.1 mrg price. Our General Public Licenses are designed to make sure that you
3176 1.1 mrg have the freedom to distribute copies of free software (and charge for
3177 1.3 mrg them if you wish), that you receive source code or can get it if you
3178 1.3 mrg want it, that you can change the software or use pieces of it in new
3179 1.3 mrg free programs, and that you know you can do these things.
3180 1.3 mrg
3181 1.3 mrg To protect your rights, we need to prevent others from denying you
3182 1.11 mrg these rights or asking you to surrender the rights. Therefore, you have
3183 1.11 mrg certain responsibilities if you distribute copies of the software, or if
3184 1.11 mrg you modify it: responsibilities to respect the freedom of others.
3185 1.1 mrg
3186 1.1 mrg For example, if you distribute copies of such a program, whether
3187 1.3 mrg gratis or for a fee, you must pass on to the recipients the same
3188 1.3 mrg freedoms that you received. You must make sure that they, too, receive
3189 1.3 mrg or can get the source code. And you must show them these terms so they
3190 1.3 mrg know their rights.
3191 1.3 mrg
3192 1.3 mrg Developers that use the GNU GPL protect your rights with two steps:
3193 1.3 mrg (1) assert copyright on the software, and (2) offer you this License
3194 1.3 mrg giving you legal permission to copy, distribute and/or modify it.
3195 1.3 mrg
3196 1.3 mrg For the developers' and authors' protection, the GPL clearly explains
3197 1.3 mrg that there is no warranty for this free software. For both users' and
3198 1.3 mrg authors' sake, the GPL requires that modified versions be marked as
3199 1.3 mrg changed, so that their problems will not be attributed erroneously to
3200 1.3 mrg authors of previous versions.
3201 1.3 mrg
3202 1.3 mrg Some devices are designed to deny users access to install or run
3203 1.11 mrg modified versions of the software inside them, although the manufacturer
3204 1.11 mrg can do so. This is fundamentally incompatible with the aim of
3205 1.11 mrg protecting users' freedom to change the software. The systematic
3206 1.11 mrg pattern of such abuse occurs in the area of products for individuals to
3207 1.11 mrg use, which is precisely where it is most unacceptable. Therefore, we
3208 1.11 mrg have designed this version of the GPL to prohibit the practice for those
3209 1.11 mrg products. If such problems arise substantially in other domains, we
3210 1.11 mrg stand ready to extend this provision to those domains in future versions
3211 1.11 mrg of the GPL, as needed to protect the freedom of users.
3212 1.3 mrg
3213 1.3 mrg Finally, every program is threatened constantly by software patents.
3214 1.3 mrg States should not allow patents to restrict development and use of
3215 1.3 mrg software on general-purpose computers, but in those that do, we wish to
3216 1.3 mrg avoid the special danger that patents applied to a free program could
3217 1.3 mrg make it effectively proprietary. To prevent this, the GPL assures that
3218 1.3 mrg patents cannot be used to render the program non-free.
3219 1.1 mrg
3220 1.1 mrg The precise terms and conditions for copying, distribution and
3221 1.1 mrg modification follow.
3222 1.1 mrg
3223 1.3 mrg TERMS AND CONDITIONS
3224 1.3 mrg ====================
3225 1.3 mrg
3226 1.3 mrg 0. Definitions.
3227 1.3 mrg
3228 1.3 mrg "This License" refers to version 3 of the GNU General Public
3229 1.3 mrg License.
3230 1.1 mrg
3231 1.3 mrg "Copyright" also means copyright-like laws that apply to other
3232 1.3 mrg kinds of works, such as semiconductor masks.
3233 1.3 mrg
3234 1.3 mrg "The Program" refers to any copyrightable work licensed under this
3235 1.3 mrg License. Each licensee is addressed as "you". "Licensees" and
3236 1.3 mrg "recipients" may be individuals or organizations.
3237 1.3 mrg
3238 1.3 mrg To "modify" a work means to copy from or adapt all or part of the
3239 1.3 mrg work in a fashion requiring copyright permission, other than the
3240 1.3 mrg making of an exact copy. The resulting work is called a "modified
3241 1.3 mrg version" of the earlier work or a work "based on" the earlier work.
3242 1.3 mrg
3243 1.3 mrg A "covered work" means either the unmodified Program or a work
3244 1.3 mrg based on the Program.
3245 1.3 mrg
3246 1.3 mrg To "propagate" a work means to do anything with it that, without
3247 1.3 mrg permission, would make you directly or secondarily liable for
3248 1.11 mrg infringement under applicable copyright law, except executing it on
3249 1.11 mrg a computer or modifying a private copy. Propagation includes
3250 1.3 mrg copying, distribution (with or without modification), making
3251 1.3 mrg available to the public, and in some countries other activities as
3252 1.3 mrg well.
3253 1.3 mrg
3254 1.3 mrg To "convey" a work means any kind of propagation that enables other
3255 1.3 mrg parties to make or receive copies. Mere interaction with a user
3256 1.3 mrg through a computer network, with no transfer of a copy, is not
3257 1.3 mrg conveying.
3258 1.3 mrg
3259 1.3 mrg An interactive user interface displays "Appropriate Legal Notices"
3260 1.3 mrg to the extent that it includes a convenient and prominently visible
3261 1.3 mrg feature that (1) displays an appropriate copyright notice, and (2)
3262 1.3 mrg tells the user that there is no warranty for the work (except to
3263 1.11 mrg the extent that warranties are provided), that licensees may convey
3264 1.11 mrg the work under this License, and how to view a copy of this
3265 1.3 mrg License. If the interface presents a list of user commands or
3266 1.3 mrg options, such as a menu, a prominent item in the list meets this
3267 1.3 mrg criterion.
3268 1.3 mrg
3269 1.3 mrg 1. Source Code.
3270 1.3 mrg
3271 1.3 mrg The "source code" for a work means the preferred form of the work
3272 1.11 mrg for making modifications to it. "Object code" means any non-source
3273 1.11 mrg form of a work.
3274 1.3 mrg
3275 1.3 mrg A "Standard Interface" means an interface that either is an
3276 1.3 mrg official standard defined by a recognized standards body, or, in
3277 1.3 mrg the case of interfaces specified for a particular programming
3278 1.3 mrg language, one that is widely used among developers working in that
3279 1.3 mrg language.
3280 1.3 mrg
3281 1.3 mrg The "System Libraries" of an executable work include anything,
3282 1.3 mrg other than the work as a whole, that (a) is included in the normal
3283 1.3 mrg form of packaging a Major Component, but which is not part of that
3284 1.11 mrg Major Component, and (b) serves only to enable use of the work with
3285 1.11 mrg that Major Component, or to implement a Standard Interface for
3286 1.11 mrg which an implementation is available to the public in source code
3287 1.11 mrg form. A "Major Component", in this context, means a major
3288 1.3 mrg essential component (kernel, window system, and so on) of the
3289 1.3 mrg specific operating system (if any) on which the executable work
3290 1.3 mrg runs, or a compiler used to produce the work, or an object code
3291 1.3 mrg interpreter used to run it.
3292 1.3 mrg
3293 1.3 mrg The "Corresponding Source" for a work in object code form means all
3294 1.3 mrg the source code needed to generate, install, and (for an executable
3295 1.11 mrg work) run the object code and to modify the work, including scripts
3296 1.11 mrg to control those activities. However, it does not include the
3297 1.11 mrg work's System Libraries, or general-purpose tools or generally
3298 1.3 mrg available free programs which are used unmodified in performing
3299 1.3 mrg those activities but which are not part of the work. For example,
3300 1.11 mrg Corresponding Source includes interface definition files associated
3301 1.11 mrg with source files for the work, and the source code for shared
3302 1.11 mrg libraries and dynamically linked subprograms that the work is
3303 1.11 mrg specifically designed to require, such as by intimate data
3304 1.3 mrg communication or control flow between those subprograms and other
3305 1.3 mrg parts of the work.
3306 1.3 mrg
3307 1.3 mrg The Corresponding Source need not include anything that users can
3308 1.3 mrg regenerate automatically from other parts of the Corresponding
3309 1.3 mrg Source.
3310 1.3 mrg
3311 1.3 mrg The Corresponding Source for a work in source code form is that
3312 1.3 mrg same work.
3313 1.3 mrg
3314 1.3 mrg 2. Basic Permissions.
3315 1.3 mrg
3316 1.3 mrg All rights granted under this License are granted for the term of
3317 1.3 mrg copyright on the Program, and are irrevocable provided the stated
3318 1.3 mrg conditions are met. This License explicitly affirms your unlimited
3319 1.3 mrg permission to run the unmodified Program. The output from running
3320 1.11 mrg a covered work is covered by this License only if the output, given
3321 1.11 mrg its content, constitutes a covered work. This License acknowledges
3322 1.11 mrg your rights of fair use or other equivalent, as provided by
3323 1.11 mrg copyright law.
3324 1.3 mrg
3325 1.3 mrg You may make, run and propagate covered works that you do not
3326 1.3 mrg convey, without conditions so long as your license otherwise
3327 1.3 mrg remains in force. You may convey covered works to others for the
3328 1.11 mrg sole purpose of having them make modifications exclusively for you,
3329 1.11 mrg or provide you with facilities for running those works, provided
3330 1.11 mrg that you comply with the terms of this License in conveying all
3331 1.11 mrg material for which you do not control copyright. Those thus making
3332 1.11 mrg or running the covered works for you must do so exclusively on your
3333 1.11 mrg behalf, under your direction and control, on terms that prohibit
3334 1.11 mrg them from making any copies of your copyrighted material outside
3335 1.11 mrg their relationship with you.
3336 1.3 mrg
3337 1.3 mrg Conveying under any other circumstances is permitted solely under
3338 1.3 mrg the conditions stated below. Sublicensing is not allowed; section
3339 1.3 mrg 10 makes it unnecessary.
3340 1.3 mrg
3341 1.3 mrg 3. Protecting Users' Legal Rights From Anti-Circumvention Law.
3342 1.3 mrg
3343 1.3 mrg No covered work shall be deemed part of an effective technological
3344 1.3 mrg measure under any applicable law fulfilling obligations under
3345 1.3 mrg article 11 of the WIPO copyright treaty adopted on 20 December
3346 1.3 mrg 1996, or similar laws prohibiting or restricting circumvention of
3347 1.3 mrg such measures.
3348 1.3 mrg
3349 1.3 mrg When you convey a covered work, you waive any legal power to forbid
3350 1.3 mrg circumvention of technological measures to the extent such
3351 1.3 mrg circumvention is effected by exercising rights under this License
3352 1.11 mrg with respect to the covered work, and you disclaim any intention to
3353 1.11 mrg limit operation or modification of the work as a means of
3354 1.3 mrg enforcing, against the work's users, your or third parties' legal
3355 1.3 mrg rights to forbid circumvention of technological measures.
3356 1.3 mrg
3357 1.3 mrg 4. Conveying Verbatim Copies.
3358 1.3 mrg
3359 1.3 mrg You may convey verbatim copies of the Program's source code as you
3360 1.3 mrg receive it, in any medium, provided that you conspicuously and
3361 1.3 mrg appropriately publish on each copy an appropriate copyright notice;
3362 1.3 mrg keep intact all notices stating that this License and any
3363 1.3 mrg non-permissive terms added in accord with section 7 apply to the
3364 1.3 mrg code; keep intact all notices of the absence of any warranty; and
3365 1.3 mrg give all recipients a copy of this License along with the Program.
3366 1.3 mrg
3367 1.3 mrg You may charge any price or no price for each copy that you convey,
3368 1.3 mrg and you may offer support or warranty protection for a fee.
3369 1.3 mrg
3370 1.3 mrg 5. Conveying Modified Source Versions.
3371 1.3 mrg
3372 1.3 mrg You may convey a work based on the Program, or the modifications to
3373 1.3 mrg produce it from the Program, in the form of source code under the
3374 1.3 mrg terms of section 4, provided that you also meet all of these
3375 1.3 mrg conditions:
3376 1.3 mrg
3377 1.3 mrg a. The work must carry prominent notices stating that you
3378 1.3 mrg modified it, and giving a relevant date.
3379 1.3 mrg
3380 1.3 mrg b. The work must carry prominent notices stating that it is
3381 1.3 mrg released under this License and any conditions added under
3382 1.3 mrg section 7. This requirement modifies the requirement in
3383 1.3 mrg section 4 to "keep intact all notices".
3384 1.3 mrg
3385 1.3 mrg c. You must license the entire work, as a whole, under this
3386 1.3 mrg License to anyone who comes into possession of a copy. This
3387 1.3 mrg License will therefore apply, along with any applicable
3388 1.3 mrg section 7 additional terms, to the whole of the work, and all
3389 1.3 mrg its parts, regardless of how they are packaged. This License
3390 1.3 mrg gives no permission to license the work in any other way, but
3391 1.3 mrg it does not invalidate such permission if you have separately
3392 1.3 mrg received it.
3393 1.3 mrg
3394 1.3 mrg d. If the work has interactive user interfaces, each must display
3395 1.3 mrg Appropriate Legal Notices; however, if the Program has
3396 1.3 mrg interactive interfaces that do not display Appropriate Legal
3397 1.3 mrg Notices, your work need not make them do so.
3398 1.3 mrg
3399 1.3 mrg A compilation of a covered work with other separate and independent
3400 1.3 mrg works, which are not by their nature extensions of the covered
3401 1.3 mrg work, and which are not combined with it such as to form a larger
3402 1.3 mrg program, in or on a volume of a storage or distribution medium, is
3403 1.3 mrg called an "aggregate" if the compilation and its resulting
3404 1.3 mrg copyright are not used to limit the access or legal rights of the
3405 1.3 mrg compilation's users beyond what the individual works permit.
3406 1.3 mrg Inclusion of a covered work in an aggregate does not cause this
3407 1.3 mrg License to apply to the other parts of the aggregate.
3408 1.3 mrg
3409 1.3 mrg 6. Conveying Non-Source Forms.
3410 1.3 mrg
3411 1.3 mrg You may convey a covered work in object code form under the terms
3412 1.3 mrg of sections 4 and 5, provided that you also convey the
3413 1.3 mrg machine-readable Corresponding Source under the terms of this
3414 1.3 mrg License, in one of these ways:
3415 1.3 mrg
3416 1.3 mrg a. Convey the object code in, or embodied in, a physical product
3417 1.3 mrg (including a physical distribution medium), accompanied by the
3418 1.3 mrg Corresponding Source fixed on a durable physical medium
3419 1.3 mrg customarily used for software interchange.
3420 1.3 mrg
3421 1.3 mrg b. Convey the object code in, or embodied in, a physical product
3422 1.3 mrg (including a physical distribution medium), accompanied by a
3423 1.11 mrg written offer, valid for at least three years and valid for as
3424 1.11 mrg long as you offer spare parts or customer support for that
3425 1.3 mrg product model, to give anyone who possesses the object code
3426 1.3 mrg either (1) a copy of the Corresponding Source for all the
3427 1.3 mrg software in the product that is covered by this License, on a
3428 1.3 mrg durable physical medium customarily used for software
3429 1.3 mrg interchange, for a price no more than your reasonable cost of
3430 1.3 mrg physically performing this conveying of source, or (2) access
3431 1.3 mrg to copy the Corresponding Source from a network server at no
3432 1.3 mrg charge.
3433 1.3 mrg
3434 1.11 mrg c. Convey individual copies of the object code with a copy of the
3435 1.11 mrg written offer to provide the Corresponding Source. This
3436 1.3 mrg alternative is allowed only occasionally and noncommercially,
3437 1.3 mrg and only if you received the object code with such an offer,
3438 1.3 mrg in accord with subsection 6b.
3439 1.3 mrg
3440 1.3 mrg d. Convey the object code by offering access from a designated
3441 1.11 mrg place (gratis or for a charge), and offer equivalent access to
3442 1.11 mrg the Corresponding Source in the same way through the same
3443 1.3 mrg place at no further charge. You need not require recipients
3444 1.3 mrg to copy the Corresponding Source along with the object code.
3445 1.3 mrg If the place to copy the object code is a network server, the
3446 1.11 mrg Corresponding Source may be on a different server (operated by
3447 1.11 mrg you or a third party) that supports equivalent copying
3448 1.11 mrg facilities, provided you maintain clear directions next to the
3449 1.11 mrg object code saying where to find the Corresponding Source.
3450 1.3 mrg Regardless of what server hosts the Corresponding Source, you
3451 1.11 mrg remain obligated to ensure that it is available for as long as
3452 1.11 mrg needed to satisfy these requirements.
3453 1.3 mrg
3454 1.3 mrg e. Convey the object code using peer-to-peer transmission,
3455 1.3 mrg provided you inform other peers where the object code and
3456 1.3 mrg Corresponding Source of the work are being offered to the
3457 1.3 mrg general public at no charge under subsection 6d.
3458 1.3 mrg
3459 1.3 mrg A separable portion of the object code, whose source code is
3460 1.3 mrg excluded from the Corresponding Source as a System Library, need
3461 1.3 mrg not be included in conveying the object code work.
3462 1.3 mrg
3463 1.3 mrg A "User Product" is either (1) a "consumer product", which means
3464 1.3 mrg any tangible personal property which is normally used for personal,
3465 1.3 mrg family, or household purposes, or (2) anything designed or sold for
3466 1.11 mrg incorporation into a dwelling. In determining whether a product is
3467 1.11 mrg a consumer product, doubtful cases shall be resolved in favor of
3468 1.3 mrg coverage. For a particular product received by a particular user,
3469 1.3 mrg "normally used" refers to a typical or common use of that class of
3470 1.3 mrg product, regardless of the status of the particular user or of the
3471 1.3 mrg way in which the particular user actually uses, or expects or is
3472 1.3 mrg expected to use, the product. A product is a consumer product
3473 1.3 mrg regardless of whether the product has substantial commercial,
3474 1.3 mrg industrial or non-consumer uses, unless such uses represent the
3475 1.3 mrg only significant mode of use of the product.
3476 1.3 mrg
3477 1.3 mrg "Installation Information" for a User Product means any methods,
3478 1.3 mrg procedures, authorization keys, or other information required to
3479 1.3 mrg install and execute modified versions of a covered work in that
3480 1.3 mrg User Product from a modified version of its Corresponding Source.
3481 1.3 mrg The information must suffice to ensure that the continued
3482 1.3 mrg functioning of the modified object code is in no case prevented or
3483 1.3 mrg interfered with solely because modification has been made.
3484 1.3 mrg
3485 1.3 mrg If you convey an object code work under this section in, or with,
3486 1.3 mrg or specifically for use in, a User Product, and the conveying
3487 1.3 mrg occurs as part of a transaction in which the right of possession
3488 1.3 mrg and use of the User Product is transferred to the recipient in
3489 1.3 mrg perpetuity or for a fixed term (regardless of how the transaction
3490 1.3 mrg is characterized), the Corresponding Source conveyed under this
3491 1.3 mrg section must be accompanied by the Installation Information. But
3492 1.3 mrg this requirement does not apply if neither you nor any third party
3493 1.3 mrg retains the ability to install modified object code on the User
3494 1.3 mrg Product (for example, the work has been installed in ROM).
3495 1.3 mrg
3496 1.3 mrg The requirement to provide Installation Information does not
3497 1.3 mrg include a requirement to continue to provide support service,
3498 1.11 mrg warranty, or updates for a work that has been modified or installed
3499 1.11 mrg by the recipient, or for the User Product in which it has been
3500 1.11 mrg modified or installed. Access to a network may be denied when the
3501 1.11 mrg modification itself materially and adversely affects the operation
3502 1.11 mrg of the network or violates the rules and protocols for
3503 1.3 mrg communication across the network.
3504 1.3 mrg
3505 1.3 mrg Corresponding Source conveyed, and Installation Information
3506 1.3 mrg provided, in accord with this section must be in a format that is
3507 1.3 mrg publicly documented (and with an implementation available to the
3508 1.3 mrg public in source code form), and must require no special password
3509 1.3 mrg or key for unpacking, reading or copying.
3510 1.3 mrg
3511 1.3 mrg 7. Additional Terms.
3512 1.3 mrg
3513 1.3 mrg "Additional permissions" are terms that supplement the terms of
3514 1.3 mrg this License by making exceptions from one or more of its
3515 1.3 mrg conditions. Additional permissions that are applicable to the
3516 1.3 mrg entire Program shall be treated as though they were included in
3517 1.3 mrg this License, to the extent that they are valid under applicable
3518 1.3 mrg law. If additional permissions apply only to part of the Program,
3519 1.3 mrg that part may be used separately under those permissions, but the
3520 1.3 mrg entire Program remains governed by this License without regard to
3521 1.3 mrg the additional permissions.
3522 1.3 mrg
3523 1.3 mrg When you convey a copy of a covered work, you may at your option
3524 1.3 mrg remove any additional permissions from that copy, or from any part
3525 1.3 mrg of it. (Additional permissions may be written to require their own
3526 1.3 mrg removal in certain cases when you modify the work.) You may place
3527 1.3 mrg additional permissions on material, added by you to a covered work,
3528 1.3 mrg for which you have or can give appropriate copyright permission.
3529 1.3 mrg
3530 1.3 mrg Notwithstanding any other provision of this License, for material
3531 1.3 mrg you add to a covered work, you may (if authorized by the copyright
3532 1.11 mrg holders of that material) supplement the terms of this License with
3533 1.11 mrg terms:
3534 1.3 mrg
3535 1.3 mrg a. Disclaiming warranty or limiting liability differently from
3536 1.3 mrg the terms of sections 15 and 16 of this License; or
3537 1.3 mrg
3538 1.3 mrg b. Requiring preservation of specified reasonable legal notices
3539 1.3 mrg or author attributions in that material or in the Appropriate
3540 1.3 mrg Legal Notices displayed by works containing it; or
3541 1.3 mrg
3542 1.3 mrg c. Prohibiting misrepresentation of the origin of that material,
3543 1.11 mrg or requiring that modified versions of such material be marked
3544 1.11 mrg in reasonable ways as different from the original version; or
3545 1.3 mrg
3546 1.3 mrg d. Limiting the use for publicity purposes of names of licensors
3547 1.3 mrg or authors of the material; or
3548 1.3 mrg
3549 1.3 mrg e. Declining to grant rights under trademark law for use of some
3550 1.3 mrg trade names, trademarks, or service marks; or
3551 1.3 mrg
3552 1.3 mrg f. Requiring indemnification of licensors and authors of that
3553 1.3 mrg material by anyone who conveys the material (or modified
3554 1.3 mrg versions of it) with contractual assumptions of liability to
3555 1.3 mrg the recipient, for any liability that these contractual
3556 1.3 mrg assumptions directly impose on those licensors and authors.
3557 1.3 mrg
3558 1.3 mrg All other non-permissive additional terms are considered "further
3559 1.3 mrg restrictions" within the meaning of section 10. If the Program as
3560 1.3 mrg you received it, or any part of it, contains a notice stating that
3561 1.3 mrg it is governed by this License along with a term that is a further
3562 1.3 mrg restriction, you may remove that term. If a license document
3563 1.11 mrg contains a further restriction but permits relicensing or conveying
3564 1.11 mrg under this License, you may add to a covered work material governed
3565 1.11 mrg by the terms of that license document, provided that the further
3566 1.11 mrg restriction does not survive such relicensing or conveying.
3567 1.3 mrg
3568 1.3 mrg If you add terms to a covered work in accord with this section, you
3569 1.3 mrg must place, in the relevant source files, a statement of the
3570 1.3 mrg additional terms that apply to those files, or a notice indicating
3571 1.3 mrg where to find the applicable terms.
3572 1.3 mrg
3573 1.3 mrg Additional terms, permissive or non-permissive, may be stated in
3574 1.3 mrg the form of a separately written license, or stated as exceptions;
3575 1.3 mrg the above requirements apply either way.
3576 1.3 mrg
3577 1.3 mrg 8. Termination.
3578 1.3 mrg
3579 1.3 mrg You may not propagate or modify a covered work except as expressly
3580 1.3 mrg provided under this License. Any attempt otherwise to propagate or
3581 1.3 mrg modify it is void, and will automatically terminate your rights
3582 1.11 mrg under this License (including any patent licenses granted under the
3583 1.11 mrg third paragraph of section 11).
3584 1.3 mrg
3585 1.3 mrg However, if you cease all violation of this License, then your
3586 1.3 mrg license from a particular copyright holder is reinstated (a)
3587 1.11 mrg provisionally, unless and until the copyright holder explicitly and
3588 1.11 mrg finally terminates your license, and (b) permanently, if the
3589 1.3 mrg copyright holder fails to notify you of the violation by some
3590 1.3 mrg reasonable means prior to 60 days after the cessation.
3591 1.3 mrg
3592 1.3 mrg Moreover, your license from a particular copyright holder is
3593 1.3 mrg reinstated permanently if the copyright holder notifies you of the
3594 1.3 mrg violation by some reasonable means, this is the first time you have
3595 1.3 mrg received notice of violation of this License (for any work) from
3596 1.3 mrg that copyright holder, and you cure the violation prior to 30 days
3597 1.3 mrg after your receipt of the notice.
3598 1.3 mrg
3599 1.3 mrg Termination of your rights under this section does not terminate
3600 1.11 mrg the licenses of parties who have received copies or rights from you
3601 1.11 mrg under this License. If your rights have been terminated and not
3602 1.11 mrg permanently reinstated, you do not qualify to receive new licenses
3603 1.11 mrg for the same material under section 10.
3604 1.3 mrg
3605 1.3 mrg 9. Acceptance Not Required for Having Copies.
3606 1.3 mrg
3607 1.3 mrg You are not required to accept this License in order to receive or
3608 1.3 mrg run a copy of the Program. Ancillary propagation of a covered work
3609 1.3 mrg occurring solely as a consequence of using peer-to-peer
3610 1.3 mrg transmission to receive a copy likewise does not require
3611 1.3 mrg acceptance. However, nothing other than this License grants you
3612 1.3 mrg permission to propagate or modify any covered work. These actions
3613 1.3 mrg infringe copyright if you do not accept this License. Therefore,
3614 1.3 mrg by modifying or propagating a covered work, you indicate your
3615 1.3 mrg acceptance of this License to do so.
3616 1.3 mrg
3617 1.11 mrg 10. Automatic Licensing of Downstream Recipients.
3618 1.3 mrg
3619 1.3 mrg Each time you convey a covered work, the recipient automatically
3620 1.3 mrg receives a license from the original licensors, to run, modify and
3621 1.3 mrg propagate that work, subject to this License. You are not
3622 1.3 mrg responsible for enforcing compliance by third parties with this
3623 1.3 mrg License.
3624 1.3 mrg
3625 1.3 mrg An "entity transaction" is a transaction transferring control of an
3626 1.3 mrg organization, or substantially all assets of one, or subdividing an
3627 1.3 mrg organization, or merging organizations. If propagation of a
3628 1.3 mrg covered work results from an entity transaction, each party to that
3629 1.3 mrg transaction who receives a copy of the work also receives whatever
3630 1.3 mrg licenses to the work the party's predecessor in interest had or
3631 1.11 mrg could give under the previous paragraph, plus a right to possession
3632 1.11 mrg of the Corresponding Source of the work from the predecessor in
3633 1.11 mrg interest, if the predecessor has it or can get it with reasonable
3634 1.11 mrg efforts.
3635 1.3 mrg
3636 1.3 mrg You may not impose any further restrictions on the exercise of the
3637 1.3 mrg rights granted or affirmed under this License. For example, you
3638 1.11 mrg may not impose a license fee, royalty, or other charge for exercise
3639 1.11 mrg of rights granted under this License, and you may not initiate
3640 1.11 mrg litigation (including a cross-claim or counterclaim in a lawsuit)
3641 1.11 mrg alleging that any patent claim is infringed by making, using,
3642 1.11 mrg selling, offering for sale, or importing the Program or any portion
3643 1.11 mrg of it.
3644 1.3 mrg
3645 1.11 mrg 11. Patents.
3646 1.3 mrg
3647 1.3 mrg A "contributor" is a copyright holder who authorizes use under this
3648 1.3 mrg License of the Program or a work on which the Program is based.
3649 1.3 mrg The work thus licensed is called the contributor's "contributor
3650 1.3 mrg version".
3651 1.3 mrg
3652 1.3 mrg A contributor's "essential patent claims" are all patent claims
3653 1.3 mrg owned or controlled by the contributor, whether already acquired or
3654 1.3 mrg hereafter acquired, that would be infringed by some manner,
3655 1.3 mrg permitted by this License, of making, using, or selling its
3656 1.3 mrg contributor version, but do not include claims that would be
3657 1.3 mrg infringed only as a consequence of further modification of the
3658 1.3 mrg contributor version. For purposes of this definition, "control"
3659 1.3 mrg includes the right to grant patent sublicenses in a manner
3660 1.3 mrg consistent with the requirements of this License.
3661 1.3 mrg
3662 1.3 mrg Each contributor grants you a non-exclusive, worldwide,
3663 1.3 mrg royalty-free patent license under the contributor's essential
3664 1.3 mrg patent claims, to make, use, sell, offer for sale, import and
3665 1.11 mrg otherwise run, modify and propagate the contents of its contributor
3666 1.11 mrg version.
3667 1.3 mrg
3668 1.3 mrg In the following three paragraphs, a "patent license" is any
3669 1.3 mrg express agreement or commitment, however denominated, not to
3670 1.3 mrg enforce a patent (such as an express permission to practice a
3671 1.11 mrg patent or covenant not to sue for patent infringement). To "grant"
3672 1.11 mrg such a patent license to a party means to make such an agreement or
3673 1.11 mrg commitment not to enforce a patent against the party.
3674 1.3 mrg
3675 1.3 mrg If you convey a covered work, knowingly relying on a patent
3676 1.3 mrg license, and the Corresponding Source of the work is not available
3677 1.3 mrg for anyone to copy, free of charge and under the terms of this
3678 1.3 mrg License, through a publicly available network server or other
3679 1.3 mrg readily accessible means, then you must either (1) cause the
3680 1.3 mrg Corresponding Source to be so available, or (2) arrange to deprive
3681 1.3 mrg yourself of the benefit of the patent license for this particular
3682 1.3 mrg work, or (3) arrange, in a manner consistent with the requirements
3683 1.3 mrg of this License, to extend the patent license to downstream
3684 1.3 mrg recipients. "Knowingly relying" means you have actual knowledge
3685 1.3 mrg that, but for the patent license, your conveying the covered work
3686 1.3 mrg in a country, or your recipient's use of the covered work in a
3687 1.3 mrg country, would infringe one or more identifiable patents in that
3688 1.3 mrg country that you have reason to believe are valid.
3689 1.3 mrg
3690 1.3 mrg If, pursuant to or in connection with a single transaction or
3691 1.3 mrg arrangement, you convey, or propagate by procuring conveyance of, a
3692 1.3 mrg covered work, and grant a patent license to some of the parties
3693 1.3 mrg receiving the covered work authorizing them to use, propagate,
3694 1.3 mrg modify or convey a specific copy of the covered work, then the
3695 1.3 mrg patent license you grant is automatically extended to all
3696 1.3 mrg recipients of the covered work and works based on it.
3697 1.3 mrg
3698 1.3 mrg A patent license is "discriminatory" if it does not include within
3699 1.3 mrg the scope of its coverage, prohibits the exercise of, or is
3700 1.3 mrg conditioned on the non-exercise of one or more of the rights that
3701 1.3 mrg are specifically granted under this License. You may not convey a
3702 1.3 mrg covered work if you are a party to an arrangement with a third
3703 1.11 mrg party that is in the business of distributing software, under which
3704 1.11 mrg you make payment to the third party based on the extent of your
3705 1.11 mrg activity of conveying the work, and under which the third party
3706 1.11 mrg grants, to any of the parties who would receive the covered work
3707 1.11 mrg from you, a discriminatory patent license (a) in connection with
3708 1.11 mrg copies of the covered work conveyed by you (or copies made from
3709 1.11 mrg those copies), or (b) primarily for and in connection with specific
3710 1.11 mrg products or compilations that contain the covered work, unless you
3711 1.11 mrg entered into that arrangement, or that patent license was granted,
3712 1.11 mrg prior to 28 March 2007.
3713 1.3 mrg
3714 1.3 mrg Nothing in this License shall be construed as excluding or limiting
3715 1.3 mrg any implied license or other defenses to infringement that may
3716 1.3 mrg otherwise be available to you under applicable patent law.
3717 1.3 mrg
3718 1.11 mrg 12. No Surrender of Others' Freedom.
3719 1.3 mrg
3720 1.11 mrg If conditions are imposed on you (whether by court order, agreement
3721 1.11 mrg or otherwise) that contradict the conditions of this License, they
3722 1.11 mrg do not excuse you from the conditions of this License. If you
3723 1.11 mrg cannot convey a covered work so as to satisfy simultaneously your
3724 1.11 mrg obligations under this License and any other pertinent obligations,
3725 1.11 mrg then as a consequence you may not convey it at all. For example,
3726 1.11 mrg if you agree to terms that obligate you to collect a royalty for
3727 1.11 mrg further conveying from those to whom you convey the Program, the
3728 1.11 mrg only way you could satisfy both those terms and this License would
3729 1.11 mrg be to refrain entirely from conveying the Program.
3730 1.3 mrg
3731 1.11 mrg 13. Use with the GNU Affero General Public License.
3732 1.3 mrg
3733 1.3 mrg Notwithstanding any other provision of this License, you have
3734 1.3 mrg permission to link or combine any covered work with a work licensed
3735 1.3 mrg under version 3 of the GNU Affero General Public License into a
3736 1.3 mrg single combined work, and to convey the resulting work. The terms
3737 1.3 mrg of this License will continue to apply to the part which is the
3738 1.3 mrg covered work, but the special requirements of the GNU Affero
3739 1.3 mrg General Public License, section 13, concerning interaction through
3740 1.3 mrg a network will apply to the combination as such.
3741 1.3 mrg
3742 1.11 mrg 14. Revised Versions of this License.
3743 1.3 mrg
3744 1.3 mrg The Free Software Foundation may publish revised and/or new
3745 1.11 mrg versions of the GNU General Public License from time to time. Such
3746 1.11 mrg new versions will be similar in spirit to the present version, but
3747 1.11 mrg may differ in detail to address new problems or concerns.
3748 1.1 mrg
3749 1.1 mrg Each version is given a distinguishing version number. If the
3750 1.3 mrg Program specifies that a certain numbered version of the GNU
3751 1.3 mrg General Public License "or any later version" applies to it, you
3752 1.3 mrg have the option of following the terms and conditions either of
3753 1.11 mrg that numbered version or of any later version published by the Free
3754 1.11 mrg Software Foundation. If the Program does not specify a version
3755 1.11 mrg number of the GNU General Public License, you may choose any
3756 1.11 mrg version ever published by the Free Software Foundation.
3757 1.1 mrg
3758 1.3 mrg If the Program specifies that a proxy can decide which future
3759 1.3 mrg versions of the GNU General Public License can be used, that
3760 1.3 mrg proxy's public statement of acceptance of a version permanently
3761 1.3 mrg authorizes you to choose that version for the Program.
3762 1.3 mrg
3763 1.3 mrg Later license versions may give you additional or different
3764 1.3 mrg permissions. However, no additional obligations are imposed on any
3765 1.3 mrg author or copyright holder as a result of your choosing to follow a
3766 1.3 mrg later version.
3767 1.3 mrg
3768 1.11 mrg 15. Disclaimer of Warranty.
3769 1.3 mrg
3770 1.3 mrg THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY
3771 1.11 mrg APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE
3772 1.3 mrg COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS"
3773 1.3 mrg WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED,
3774 1.3 mrg INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
3775 1.11 mrg MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE
3776 1.3 mrg RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU.
3777 1.3 mrg SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL
3778 1.3 mrg NECESSARY SERVICING, REPAIR OR CORRECTION.
3779 1.3 mrg
3780 1.11 mrg 16. Limitation of Liability.
3781 1.3 mrg
3782 1.3 mrg IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN
3783 1.3 mrg WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES
3784 1.11 mrg AND/OR CONVEYS THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR
3785 1.11 mrg DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR
3786 1.3 mrg CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE
3787 1.3 mrg THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA
3788 1.3 mrg BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD
3789 1.3 mrg PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER
3790 1.3 mrg PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF
3791 1.3 mrg THE POSSIBILITY OF SUCH DAMAGES.
3792 1.3 mrg
3793 1.11 mrg 17. Interpretation of Sections 15 and 16.
3794 1.3 mrg
3795 1.3 mrg If the disclaimer of warranty and limitation of liability provided
3796 1.3 mrg above cannot be given local legal effect according to their terms,
3797 1.3 mrg reviewing courts shall apply local law that most closely
3798 1.3 mrg approximates an absolute waiver of all civil liability in
3799 1.3 mrg connection with the Program, unless a warranty or assumption of
3800 1.3 mrg liability accompanies a copy of the Program in return for a fee.
3801 1.1 mrg
3802 1.3 mrg END OF TERMS AND CONDITIONS
3803 1.3 mrg ===========================
3804 1.3 mrg
3805 1.3 mrg How to Apply These Terms to Your New Programs
3806 1.3 mrg =============================================
3807 1.1 mrg
3808 1.1 mrg If you develop a new program, and you want it to be of the greatest
3809 1.1 mrg possible use to the public, the best way to achieve this is to make it
3810 1.1 mrg free software which everyone can redistribute and change under these
3811 1.1 mrg terms.
3812 1.1 mrg
3813 1.1 mrg To do so, attach the following notices to the program. It is safest
3814 1.1 mrg to attach them to the start of each source file to most effectively
3815 1.3 mrg state the exclusion of warranty; and each file should have at least the
3816 1.3 mrg "copyright" line and a pointer to where the full notice is found.
3817 1.1 mrg
3818 1.1 mrg ONE LINE TO GIVE THE PROGRAM'S NAME AND A BRIEF IDEA OF WHAT IT DOES.
3819 1.3 mrg Copyright (C) YEAR NAME OF AUTHOR
3820 1.1 mrg
3821 1.3 mrg This program is free software: you can redistribute it and/or modify
3822 1.1 mrg it under the terms of the GNU General Public License as published by
3823 1.3 mrg the Free Software Foundation, either version 3 of the License, or (at
3824 1.3 mrg your option) any later version.
3825 1.1 mrg
3826 1.3 mrg This program is distributed in the hope that it will be useful, but
3827 1.3 mrg WITHOUT ANY WARRANTY; without even the implied warranty of
3828 1.3 mrg MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
3829 1.3 mrg General Public License for more details.
3830 1.1 mrg
3831 1.1 mrg You should have received a copy of the GNU General Public License
3832 1.11 mrg along with this program. If not, see <http://www.gnu.org/licenses/>.
3833 1.1 mrg
3834 1.1 mrg Also add information on how to contact you by electronic and paper
3835 1.1 mrg mail.
3836 1.1 mrg
3837 1.3 mrg If the program does terminal interaction, make it output a short
3838 1.3 mrg notice like this when it starts in an interactive mode:
3839 1.1 mrg
3840 1.3 mrg PROGRAM Copyright (C) YEAR NAME OF AUTHOR
3841 1.11 mrg This program comes with ABSOLUTELY NO WARRANTY; for details type 'show w'.
3842 1.1 mrg This is free software, and you are welcome to redistribute it
3843 1.11 mrg under certain conditions; type 'show c' for details.
3844 1.1 mrg
3845 1.11 mrg The hypothetical commands 'show w' and 'show c' should show the
3846 1.3 mrg appropriate parts of the General Public License. Of course, your
3847 1.3 mrg program's commands might be different; for a GUI interface, you would
3848 1.3 mrg use an "about box".
3849 1.1 mrg
3850 1.1 mrg You should also get your employer (if you work as a programmer) or
3851 1.3 mrg school, if any, to sign a "copyright disclaimer" for the program, if
3852 1.3 mrg necessary. For more information on this, and how to apply and follow
3853 1.11 mrg the GNU GPL, see <http://www.gnu.org/licenses/>.
3854 1.1 mrg
3855 1.3 mrg The GNU General Public License does not permit incorporating your
3856 1.1 mrg program into proprietary programs. If your program is a subroutine
3857 1.1 mrg library, you may consider it more useful to permit linking proprietary
3858 1.1 mrg applications with the library. If this is what you want to do, use the
3859 1.3 mrg GNU Lesser General Public License instead of this License. But first,
3860 1.11 mrg please read <http://www.gnu.org/philosophy/why-not-lgpl.html>.
3861 1.1 mrg
3862 1.1 mrg
3863 1.1 mrg File: libgomp.info, Node: GNU Free Documentation License, Next: Funding, Prev: Copying, Up: Top
3864 1.1 mrg
3865 1.1 mrg GNU Free Documentation License
3866 1.1 mrg ******************************
3867 1.1 mrg
3868 1.3 mrg Version 1.3, 3 November 2008
3869 1.1 mrg
3870 1.3 mrg Copyright (C) 2000, 2001, 2002, 2007, 2008 Free Software Foundation, Inc.
3871 1.11 mrg <http://fsf.org/>
3872 1.1 mrg
3873 1.1 mrg Everyone is permitted to copy and distribute verbatim copies
3874 1.1 mrg of this license document, but changing it is not allowed.
3875 1.1 mrg
3876 1.1 mrg 0. PREAMBLE
3877 1.1 mrg
3878 1.1 mrg The purpose of this License is to make a manual, textbook, or other
3879 1.1 mrg functional and useful document "free" in the sense of freedom: to
3880 1.1 mrg assure everyone the effective freedom to copy and redistribute it,
3881 1.1 mrg with or without modifying it, either commercially or
3882 1.1 mrg noncommercially. Secondarily, this License preserves for the
3883 1.1 mrg author and publisher a way to get credit for their work, while not
3884 1.1 mrg being considered responsible for modifications made by others.
3885 1.1 mrg
3886 1.1 mrg This License is a kind of "copyleft", which means that derivative
3887 1.1 mrg works of the document must themselves be free in the same sense.
3888 1.1 mrg It complements the GNU General Public License, which is a copyleft
3889 1.1 mrg license designed for free software.
3890 1.1 mrg
3891 1.1 mrg We have designed this License in order to use it for manuals for
3892 1.1 mrg free software, because free software needs free documentation: a
3893 1.1 mrg free program should come with manuals providing the same freedoms
3894 1.1 mrg that the software does. But this License is not limited to
3895 1.1 mrg software manuals; it can be used for any textual work, regardless
3896 1.11 mrg of subject matter or whether it is published as a printed book. We
3897 1.11 mrg recommend this License principally for works whose purpose is
3898 1.1 mrg instruction or reference.
3899 1.1 mrg
3900 1.1 mrg 1. APPLICABILITY AND DEFINITIONS
3901 1.1 mrg
3902 1.1 mrg This License applies to any manual or other work, in any medium,
3903 1.11 mrg that contains a notice placed by the copyright holder saying it can
3904 1.11 mrg be distributed under the terms of this License. Such a notice
3905 1.1 mrg grants a world-wide, royalty-free license, unlimited in duration,
3906 1.1 mrg to use that work under the conditions stated herein. The
3907 1.1 mrg "Document", below, refers to any such manual or work. Any member
3908 1.11 mrg of the public is a licensee, and is addressed as "you". You accept
3909 1.11 mrg the license if you copy, modify or distribute the work in a way
3910 1.11 mrg requiring permission under copyright law.
3911 1.1 mrg
3912 1.1 mrg A "Modified Version" of the Document means any work containing the
3913 1.1 mrg Document or a portion of it, either copied verbatim, or with
3914 1.1 mrg modifications and/or translated into another language.
3915 1.1 mrg
3916 1.1 mrg A "Secondary Section" is a named appendix or a front-matter section
3917 1.1 mrg of the Document that deals exclusively with the relationship of the
3918 1.1 mrg publishers or authors of the Document to the Document's overall
3919 1.1 mrg subject (or to related matters) and contains nothing that could
3920 1.1 mrg fall directly within that overall subject. (Thus, if the Document
3921 1.1 mrg is in part a textbook of mathematics, a Secondary Section may not
3922 1.1 mrg explain any mathematics.) The relationship could be a matter of
3923 1.1 mrg historical connection with the subject or with related matters, or
3924 1.1 mrg of legal, commercial, philosophical, ethical or political position
3925 1.1 mrg regarding them.
3926 1.1 mrg
3927 1.1 mrg The "Invariant Sections" are certain Secondary Sections whose
3928 1.11 mrg titles are designated, as being those of Invariant Sections, in the
3929 1.11 mrg notice that says that the Document is released under this License.
3930 1.11 mrg If a section does not fit the above definition of Secondary then it
3931 1.11 mrg is not allowed to be designated as Invariant. The Document may
3932 1.11 mrg contain zero Invariant Sections. If the Document does not identify
3933 1.11 mrg any Invariant Sections then there are none.
3934 1.1 mrg
3935 1.1 mrg The "Cover Texts" are certain short passages of text that are
3936 1.1 mrg listed, as Front-Cover Texts or Back-Cover Texts, in the notice
3937 1.1 mrg that says that the Document is released under this License. A
3938 1.1 mrg Front-Cover Text may be at most 5 words, and a Back-Cover Text may
3939 1.1 mrg be at most 25 words.
3940 1.1 mrg
3941 1.1 mrg A "Transparent" copy of the Document means a machine-readable copy,
3942 1.1 mrg represented in a format whose specification is available to the
3943 1.1 mrg general public, that is suitable for revising the document
3944 1.11 mrg straightforwardly with generic text editors or (for images composed
3945 1.11 mrg of pixels) generic paint programs or (for drawings) some widely
3946 1.11 mrg available drawing editor, and that is suitable for input to text
3947 1.11 mrg formatters or for automatic translation to a variety of formats
3948 1.11 mrg suitable for input to text formatters. A copy made in an otherwise
3949 1.11 mrg Transparent file format whose markup, or absence of markup, has
3950 1.11 mrg been arranged to thwart or discourage subsequent modification by
3951 1.11 mrg readers is not Transparent. An image format is not Transparent if
3952 1.11 mrg used for any substantial amount of text. A copy that is not
3953 1.11 mrg "Transparent" is called "Opaque".
3954 1.1 mrg
3955 1.1 mrg Examples of suitable formats for Transparent copies include plain
3956 1.1 mrg ASCII without markup, Texinfo input format, LaTeX input format,
3957 1.11 mrg SGML or XML using a publicly available DTD, and standard-conforming
3958 1.11 mrg simple HTML, PostScript or PDF designed for human modification.
3959 1.11 mrg Examples of transparent image formats include PNG, XCF and JPG.
3960 1.11 mrg Opaque formats include proprietary formats that can be read and
3961 1.11 mrg edited only by proprietary word processors, SGML or XML for which
3962 1.11 mrg the DTD and/or processing tools are not generally available, and
3963 1.11 mrg the machine-generated HTML, PostScript or PDF produced by some word
3964 1.11 mrg processors for output purposes only.
3965 1.1 mrg
3966 1.1 mrg The "Title Page" means, for a printed book, the title page itself,
3967 1.1 mrg plus such following pages as are needed to hold, legibly, the
3968 1.1 mrg material this License requires to appear in the title page. For
3969 1.1 mrg works in formats which do not have any title page as such, "Title
3970 1.1 mrg Page" means the text near the most prominent appearance of the
3971 1.1 mrg work's title, preceding the beginning of the body of the text.
3972 1.1 mrg
3973 1.3 mrg The "publisher" means any person or entity that distributes copies
3974 1.3 mrg of the Document to the public.
3975 1.3 mrg
3976 1.1 mrg A section "Entitled XYZ" means a named subunit of the Document
3977 1.1 mrg whose title either is precisely XYZ or contains XYZ in parentheses
3978 1.1 mrg following text that translates XYZ in another language. (Here XYZ
3979 1.1 mrg stands for a specific section name mentioned below, such as
3980 1.1 mrg "Acknowledgements", "Dedications", "Endorsements", or "History".)
3981 1.1 mrg To "Preserve the Title" of such a section when you modify the
3982 1.1 mrg Document means that it remains a section "Entitled XYZ" according
3983 1.1 mrg to this definition.
3984 1.1 mrg
3985 1.1 mrg The Document may include Warranty Disclaimers next to the notice
3986 1.1 mrg which states that this License applies to the Document. These
3987 1.1 mrg Warranty Disclaimers are considered to be included by reference in
3988 1.1 mrg this License, but only as regards disclaiming warranties: any other
3989 1.1 mrg implication that these Warranty Disclaimers may have is void and
3990 1.1 mrg has no effect on the meaning of this License.
3991 1.1 mrg
3992 1.1 mrg 2. VERBATIM COPYING
3993 1.1 mrg
3994 1.1 mrg You may copy and distribute the Document in any medium, either
3995 1.1 mrg commercially or noncommercially, provided that this License, the
3996 1.1 mrg copyright notices, and the license notice saying this License
3997 1.1 mrg applies to the Document are reproduced in all copies, and that you
3998 1.1 mrg add no other conditions whatsoever to those of this License. You
3999 1.1 mrg may not use technical measures to obstruct or control the reading
4000 1.1 mrg or further copying of the copies you make or distribute. However,
4001 1.1 mrg you may accept compensation in exchange for copies. If you
4002 1.11 mrg distribute a large enough number of copies you must also follow the
4003 1.11 mrg conditions in section 3.
4004 1.1 mrg
4005 1.1 mrg You may also lend copies, under the same conditions stated above,
4006 1.1 mrg and you may publicly display copies.
4007 1.1 mrg
4008 1.1 mrg 3. COPYING IN QUANTITY
4009 1.1 mrg
4010 1.1 mrg If you publish printed copies (or copies in media that commonly
4011 1.1 mrg have printed covers) of the Document, numbering more than 100, and
4012 1.1 mrg the Document's license notice requires Cover Texts, you must
4013 1.1 mrg enclose the copies in covers that carry, clearly and legibly, all
4014 1.1 mrg these Cover Texts: Front-Cover Texts on the front cover, and
4015 1.1 mrg Back-Cover Texts on the back cover. Both covers must also clearly
4016 1.1 mrg and legibly identify you as the publisher of these copies. The
4017 1.11 mrg front cover must present the full title with all words of the title
4018 1.11 mrg equally prominent and visible. You may add other material on the
4019 1.11 mrg covers in addition. Copying with changes limited to the covers, as
4020 1.11 mrg long as they preserve the title of the Document and satisfy these
4021 1.11 mrg conditions, can be treated as verbatim copying in other respects.
4022 1.1 mrg
4023 1.1 mrg If the required texts for either cover are too voluminous to fit
4024 1.1 mrg legibly, you should put the first ones listed (as many as fit
4025 1.1 mrg reasonably) on the actual cover, and continue the rest onto
4026 1.1 mrg adjacent pages.
4027 1.1 mrg
4028 1.1 mrg If you publish or distribute Opaque copies of the Document
4029 1.11 mrg numbering more than 100, you must either include a machine-readable
4030 1.11 mrg Transparent copy along with each Opaque copy, or state in or with
4031 1.11 mrg each Opaque copy a computer-network location from which the general
4032 1.11 mrg network-using public has access to download using public-standard
4033 1.11 mrg network protocols a complete Transparent copy of the Document, free
4034 1.11 mrg of added material. If you use the latter option, you must take
4035 1.11 mrg reasonably prudent steps, when you begin distribution of Opaque
4036 1.11 mrg copies in quantity, to ensure that this Transparent copy will
4037 1.11 mrg remain thus accessible at the stated location until at least one
4038 1.11 mrg year after the last time you distribute an Opaque copy (directly or
4039 1.11 mrg through your agents or retailers) of that edition to the public.
4040 1.1 mrg
4041 1.1 mrg It is requested, but not required, that you contact the authors of
4042 1.11 mrg the Document well before redistributing any large number of copies,
4043 1.11 mrg to give them a chance to provide you with an updated version of the
4044 1.11 mrg Document.
4045 1.1 mrg
4046 1.1 mrg 4. MODIFICATIONS
4047 1.1 mrg
4048 1.1 mrg You may copy and distribute a Modified Version of the Document
4049 1.1 mrg under the conditions of sections 2 and 3 above, provided that you
4050 1.11 mrg release the Modified Version under precisely this License, with the
4051 1.11 mrg Modified Version filling the role of the Document, thus licensing
4052 1.11 mrg distribution and modification of the Modified Version to whoever
4053 1.11 mrg possesses a copy of it. In addition, you must do these things in
4054 1.11 mrg the Modified Version:
4055 1.1 mrg
4056 1.1 mrg A. Use in the Title Page (and on the covers, if any) a title
4057 1.11 mrg distinct from that of the Document, and from those of previous
4058 1.11 mrg versions (which should, if there were any, be listed in the
4059 1.11 mrg History section of the Document). You may use the same title
4060 1.11 mrg as a previous version if the original publisher of that
4061 1.11 mrg version gives permission.
4062 1.1 mrg
4063 1.1 mrg B. List on the Title Page, as authors, one or more persons or
4064 1.1 mrg entities responsible for authorship of the modifications in
4065 1.1 mrg the Modified Version, together with at least five of the
4066 1.1 mrg principal authors of the Document (all of its principal
4067 1.1 mrg authors, if it has fewer than five), unless they release you
4068 1.1 mrg from this requirement.
4069 1.1 mrg
4070 1.1 mrg C. State on the Title page the name of the publisher of the
4071 1.1 mrg Modified Version, as the publisher.
4072 1.1 mrg
4073 1.1 mrg D. Preserve all the copyright notices of the Document.
4074 1.1 mrg
4075 1.1 mrg E. Add an appropriate copyright notice for your modifications
4076 1.1 mrg adjacent to the other copyright notices.
4077 1.1 mrg
4078 1.1 mrg F. Include, immediately after the copyright notices, a license
4079 1.1 mrg notice giving the public permission to use the Modified
4080 1.1 mrg Version under the terms of this License, in the form shown in
4081 1.1 mrg the Addendum below.
4082 1.1 mrg
4083 1.1 mrg G. Preserve in that license notice the full lists of Invariant
4084 1.1 mrg Sections and required Cover Texts given in the Document's
4085 1.1 mrg license notice.
4086 1.1 mrg
4087 1.1 mrg H. Include an unaltered copy of this License.
4088 1.1 mrg
4089 1.1 mrg I. Preserve the section Entitled "History", Preserve its Title,
4090 1.1 mrg and add to it an item stating at least the title, year, new
4091 1.11 mrg authors, and publisher of the Modified Version as given on the
4092 1.11 mrg Title Page. If there is no section Entitled "History" in the
4093 1.11 mrg Document, create one stating the title, year, authors, and
4094 1.11 mrg publisher of the Document as given on its Title Page, then add
4095 1.11 mrg an item describing the Modified Version as stated in the
4096 1.11 mrg previous sentence.
4097 1.1 mrg
4098 1.1 mrg J. Preserve the network location, if any, given in the Document
4099 1.1 mrg for public access to a Transparent copy of the Document, and
4100 1.1 mrg likewise the network locations given in the Document for
4101 1.11 mrg previous versions it was based on. These may be placed in the
4102 1.11 mrg "History" section. You may omit a network location for a work
4103 1.11 mrg that was published at least four years before the Document
4104 1.11 mrg itself, or if the original publisher of the version it refers
4105 1.11 mrg to gives permission.
4106 1.1 mrg
4107 1.1 mrg K. For any section Entitled "Acknowledgements" or "Dedications",
4108 1.11 mrg Preserve the Title of the section, and preserve in the section
4109 1.11 mrg all the substance and tone of each of the contributor
4110 1.1 mrg acknowledgements and/or dedications given therein.
4111 1.1 mrg
4112 1.11 mrg L. Preserve all the Invariant Sections of the Document, unaltered
4113 1.11 mrg in their text and in their titles. Section numbers or the
4114 1.11 mrg equivalent are not considered part of the section titles.
4115 1.1 mrg
4116 1.1 mrg M. Delete any section Entitled "Endorsements". Such a section
4117 1.1 mrg may not be included in the Modified Version.
4118 1.1 mrg
4119 1.1 mrg N. Do not retitle any existing section to be Entitled
4120 1.1 mrg "Endorsements" or to conflict in title with any Invariant
4121 1.1 mrg Section.
4122 1.1 mrg
4123 1.1 mrg O. Preserve any Warranty Disclaimers.
4124 1.1 mrg
4125 1.1 mrg If the Modified Version includes new front-matter sections or
4126 1.1 mrg appendices that qualify as Secondary Sections and contain no
4127 1.11 mrg material copied from the Document, you may at your option designate
4128 1.11 mrg some or all of these sections as invariant. To do this, add their
4129 1.11 mrg titles to the list of Invariant Sections in the Modified Version's
4130 1.11 mrg license notice. These titles must be distinct from any other
4131 1.11 mrg section titles.
4132 1.1 mrg
4133 1.1 mrg You may add a section Entitled "Endorsements", provided it contains
4134 1.1 mrg nothing but endorsements of your Modified Version by various
4135 1.1 mrg parties--for example, statements of peer review or that the text
4136 1.1 mrg has been approved by an organization as the authoritative
4137 1.1 mrg definition of a standard.
4138 1.1 mrg
4139 1.1 mrg You may add a passage of up to five words as a Front-Cover Text,
4140 1.11 mrg and a passage of up to 25 words as a Back-Cover Text, to the end of
4141 1.11 mrg the list of Cover Texts in the Modified Version. Only one passage
4142 1.11 mrg of Front-Cover Text and one of Back-Cover Text may be added by (or
4143 1.11 mrg through arrangements made by) any one entity. If the Document
4144 1.11 mrg already includes a cover text for the same cover, previously added
4145 1.11 mrg by you or by arrangement made by the same entity you are acting on
4146 1.11 mrg behalf of, you may not add another; but you may replace the old
4147 1.11 mrg one, on explicit permission from the previous publisher that added
4148 1.11 mrg the old one.
4149 1.1 mrg
4150 1.1 mrg The author(s) and publisher(s) of the Document do not by this
4151 1.1 mrg License give permission to use their names for publicity for or to
4152 1.1 mrg assert or imply endorsement of any Modified Version.
4153 1.1 mrg
4154 1.1 mrg 5. COMBINING DOCUMENTS
4155 1.1 mrg
4156 1.1 mrg You may combine the Document with other documents released under
4157 1.1 mrg this License, under the terms defined in section 4 above for
4158 1.11 mrg modified versions, provided that you include in the combination all
4159 1.11 mrg of the Invariant Sections of all of the original documents,
4160 1.1 mrg unmodified, and list them all as Invariant Sections of your
4161 1.1 mrg combined work in its license notice, and that you preserve all
4162 1.1 mrg their Warranty Disclaimers.
4163 1.1 mrg
4164 1.1 mrg The combined work need only contain one copy of this License, and
4165 1.1 mrg multiple identical Invariant Sections may be replaced with a single
4166 1.1 mrg copy. If there are multiple Invariant Sections with the same name
4167 1.1 mrg but different contents, make the title of each such section unique
4168 1.1 mrg by adding at the end of it, in parentheses, the name of the
4169 1.1 mrg original author or publisher of that section if known, or else a
4170 1.1 mrg unique number. Make the same adjustment to the section titles in
4171 1.1 mrg the list of Invariant Sections in the license notice of the
4172 1.1 mrg combined work.
4173 1.1 mrg
4174 1.1 mrg In the combination, you must combine any sections Entitled
4175 1.1 mrg "History" in the various original documents, forming one section
4176 1.1 mrg Entitled "History"; likewise combine any sections Entitled
4177 1.1 mrg "Acknowledgements", and any sections Entitled "Dedications". You
4178 1.1 mrg must delete all sections Entitled "Endorsements."
4179 1.1 mrg
4180 1.1 mrg 6. COLLECTIONS OF DOCUMENTS
4181 1.1 mrg
4182 1.1 mrg You may make a collection consisting of the Document and other
4183 1.1 mrg documents released under this License, and replace the individual
4184 1.1 mrg copies of this License in the various documents with a single copy
4185 1.1 mrg that is included in the collection, provided that you follow the
4186 1.11 mrg rules of this License for verbatim copying of each of the documents
4187 1.11 mrg in all other respects.
4188 1.1 mrg
4189 1.1 mrg You may extract a single document from such a collection, and
4190 1.1 mrg distribute it individually under this License, provided you insert
4191 1.11 mrg a copy of this License into the extracted document, and follow this
4192 1.11 mrg License in all other respects regarding verbatim copying of that
4193 1.11 mrg document.
4194 1.1 mrg
4195 1.1 mrg 7. AGGREGATION WITH INDEPENDENT WORKS
4196 1.1 mrg
4197 1.1 mrg A compilation of the Document or its derivatives with other
4198 1.11 mrg separate and independent documents or works, in or on a volume of a
4199 1.11 mrg storage or distribution medium, is called an "aggregate" if the
4200 1.1 mrg copyright resulting from the compilation is not used to limit the
4201 1.1 mrg legal rights of the compilation's users beyond what the individual
4202 1.1 mrg works permit. When the Document is included in an aggregate, this
4203 1.1 mrg License does not apply to the other works in the aggregate which
4204 1.1 mrg are not themselves derivative works of the Document.
4205 1.1 mrg
4206 1.1 mrg If the Cover Text requirement of section 3 is applicable to these
4207 1.1 mrg copies of the Document, then if the Document is less than one half
4208 1.1 mrg of the entire aggregate, the Document's Cover Texts may be placed
4209 1.1 mrg on covers that bracket the Document within the aggregate, or the
4210 1.1 mrg electronic equivalent of covers if the Document is in electronic
4211 1.1 mrg form. Otherwise they must appear on printed covers that bracket
4212 1.1 mrg the whole aggregate.
4213 1.1 mrg
4214 1.1 mrg 8. TRANSLATION
4215 1.1 mrg
4216 1.1 mrg Translation is considered a kind of modification, so you may
4217 1.1 mrg distribute translations of the Document under the terms of section
4218 1.1 mrg 4. Replacing Invariant Sections with translations requires special
4219 1.1 mrg permission from their copyright holders, but you may include
4220 1.1 mrg translations of some or all Invariant Sections in addition to the
4221 1.1 mrg original versions of these Invariant Sections. You may include a
4222 1.1 mrg translation of this License, and all the license notices in the
4223 1.1 mrg Document, and any Warranty Disclaimers, provided that you also
4224 1.1 mrg include the original English version of this License and the
4225 1.1 mrg original versions of those notices and disclaimers. In case of a
4226 1.1 mrg disagreement between the translation and the original version of
4227 1.1 mrg this License or a notice or disclaimer, the original version will
4228 1.1 mrg prevail.
4229 1.1 mrg
4230 1.1 mrg If a section in the Document is Entitled "Acknowledgements",
4231 1.1 mrg "Dedications", or "History", the requirement (section 4) to
4232 1.1 mrg Preserve its Title (section 1) will typically require changing the
4233 1.1 mrg actual title.
4234 1.1 mrg
4235 1.1 mrg 9. TERMINATION
4236 1.1 mrg
4237 1.1 mrg You may not copy, modify, sublicense, or distribute the Document
4238 1.3 mrg except as expressly provided under this License. Any attempt
4239 1.3 mrg otherwise to copy, modify, sublicense, or distribute it is void,
4240 1.3 mrg and will automatically terminate your rights under this License.
4241 1.3 mrg
4242 1.3 mrg However, if you cease all violation of this License, then your
4243 1.3 mrg license from a particular copyright holder is reinstated (a)
4244 1.11 mrg provisionally, unless and until the copyright holder explicitly and
4245 1.11 mrg finally terminates your license, and (b) permanently, if the
4246 1.3 mrg copyright holder fails to notify you of the violation by some
4247 1.3 mrg reasonable means prior to 60 days after the cessation.
4248 1.3 mrg
4249 1.3 mrg Moreover, your license from a particular copyright holder is
4250 1.3 mrg reinstated permanently if the copyright holder notifies you of the
4251 1.3 mrg violation by some reasonable means, this is the first time you have
4252 1.3 mrg received notice of violation of this License (for any work) from
4253 1.3 mrg that copyright holder, and you cure the violation prior to 30 days
4254 1.3 mrg after your receipt of the notice.
4255 1.3 mrg
4256 1.3 mrg Termination of your rights under this section does not terminate
4257 1.11 mrg the licenses of parties who have received copies or rights from you
4258 1.11 mrg under this License. If your rights have been terminated and not
4259 1.11 mrg permanently reinstated, receipt of a copy of some or all of the
4260 1.11 mrg same material does not give you any rights to use it.
4261 1.1 mrg
4262 1.11 mrg 10. FUTURE REVISIONS OF THIS LICENSE
4263 1.1 mrg
4264 1.1 mrg The Free Software Foundation may publish new, revised versions of
4265 1.1 mrg the GNU Free Documentation License from time to time. Such new
4266 1.1 mrg versions will be similar in spirit to the present version, but may
4267 1.1 mrg differ in detail to address new problems or concerns. See
4268 1.11 mrg <http://www.gnu.org/copyleft/>.
4269 1.1 mrg
4270 1.1 mrg Each version of the License is given a distinguishing version
4271 1.1 mrg number. If the Document specifies that a particular numbered
4272 1.1 mrg version of this License "or any later version" applies to it, you
4273 1.1 mrg have the option of following the terms and conditions either of
4274 1.1 mrg that specified version or of any later version that has been
4275 1.11 mrg published (not as a draft) by the Free Software Foundation. If the
4276 1.11 mrg Document does not specify a version number of this License, you may
4277 1.11 mrg choose any version ever published (not as a draft) by the Free
4278 1.11 mrg Software Foundation. If the Document specifies that a proxy can
4279 1.11 mrg decide which future versions of this License can be used, that
4280 1.3 mrg proxy's public statement of acceptance of a version permanently
4281 1.3 mrg authorizes you to choose that version for the Document.
4282 1.3 mrg
4283 1.11 mrg 11. RELICENSING
4284 1.3 mrg
4285 1.3 mrg "Massive Multiauthor Collaboration Site" (or "MMC Site") means any
4286 1.3 mrg World Wide Web server that publishes copyrightable works and also
4287 1.3 mrg provides prominent facilities for anybody to edit those works. A
4288 1.3 mrg public wiki that anybody can edit is an example of such a server.
4289 1.3 mrg A "Massive Multiauthor Collaboration" (or "MMC") contained in the
4290 1.3 mrg site means any set of copyrightable works thus published on the MMC
4291 1.3 mrg site.
4292 1.3 mrg
4293 1.3 mrg "CC-BY-SA" means the Creative Commons Attribution-Share Alike 3.0
4294 1.3 mrg license published by Creative Commons Corporation, a not-for-profit
4295 1.3 mrg corporation with a principal place of business in San Francisco,
4296 1.3 mrg California, as well as future copyleft versions of that license
4297 1.3 mrg published by that same organization.
4298 1.3 mrg
4299 1.3 mrg "Incorporate" means to publish or republish a Document, in whole or
4300 1.3 mrg in part, as part of another Document.
4301 1.3 mrg
4302 1.3 mrg An MMC is "eligible for relicensing" if it is licensed under this
4303 1.3 mrg License, and if all works that were first published under this
4304 1.3 mrg License somewhere other than this MMC, and subsequently
4305 1.3 mrg incorporated in whole or in part into the MMC, (1) had no cover
4306 1.3 mrg texts or invariant sections, and (2) were thus incorporated prior
4307 1.3 mrg to November 1, 2008.
4308 1.3 mrg
4309 1.3 mrg The operator of an MMC Site may republish an MMC contained in the
4310 1.3 mrg site under CC-BY-SA on the same site at any time before August 1,
4311 1.3 mrg 2009, provided the MMC is eligible for relicensing.
4312 1.3 mrg
4313 1.1 mrg ADDENDUM: How to use this License for your documents
4314 1.1 mrg ====================================================
4315 1.1 mrg
4316 1.1 mrg To use this License in a document you have written, include a copy of
4317 1.1 mrg the License in the document and put the following copyright and license
4318 1.1 mrg notices just after the title page:
4319 1.1 mrg
4320 1.1 mrg Copyright (C) YEAR YOUR NAME.
4321 1.1 mrg Permission is granted to copy, distribute and/or modify this document
4322 1.3 mrg under the terms of the GNU Free Documentation License, Version 1.3
4323 1.1 mrg or any later version published by the Free Software Foundation;
4324 1.1 mrg with no Invariant Sections, no Front-Cover Texts, and no Back-Cover
4325 1.1 mrg Texts. A copy of the license is included in the section entitled ``GNU
4326 1.1 mrg Free Documentation License''.
4327 1.1 mrg
4328 1.1 mrg If you have Invariant Sections, Front-Cover Texts and Back-Cover
4329 1.11 mrg Texts, replace the "with...Texts." line with this:
4330 1.1 mrg
4331 1.1 mrg with the Invariant Sections being LIST THEIR TITLES, with
4332 1.1 mrg the Front-Cover Texts being LIST, and with the Back-Cover Texts
4333 1.1 mrg being LIST.
4334 1.1 mrg
4335 1.1 mrg If you have Invariant Sections without Cover Texts, or some other
4336 1.1 mrg combination of the three, merge those two alternatives to suit the
4337 1.1 mrg situation.
4338 1.1 mrg
4339 1.1 mrg If your document contains nontrivial examples of program code, we
4340 1.11 mrg recommend releasing these examples in parallel under your choice of free
4341 1.11 mrg software license, such as the GNU General Public License, to permit
4342 1.11 mrg their use in free software.
4343 1.1 mrg
4344 1.1 mrg
4345 1.3 mrg File: libgomp.info, Node: Funding, Next: Library Index, Prev: GNU Free Documentation License, Up: Top
4346 1.1 mrg
4347 1.1 mrg Funding Free Software
4348 1.1 mrg *********************
4349 1.1 mrg
4350 1.1 mrg If you want to have more free software a few years from now, it makes
4351 1.1 mrg sense for you to help encourage people to contribute funds for its
4352 1.1 mrg development. The most effective approach known is to encourage
4353 1.1 mrg commercial redistributors to donate.
4354 1.1 mrg
4355 1.1 mrg Users of free software systems can boost the pace of development by
4356 1.1 mrg encouraging for-a-fee distributors to donate part of their selling price
4357 1.1 mrg to free software developers--the Free Software Foundation, and others.
4358 1.1 mrg
4359 1.1 mrg The way to convince distributors to do this is to demand it and
4360 1.1 mrg expect it from them. So when you compare distributors, judge them
4361 1.1 mrg partly by how much they give to free software development. Show
4362 1.1 mrg distributors they must compete to be the one who gives the most.
4363 1.1 mrg
4364 1.1 mrg To make this approach work, you must insist on numbers that you can
4365 1.1 mrg compare, such as, "We will donate ten dollars to the Frobnitz project
4366 1.1 mrg for each disk sold." Don't be satisfied with a vague promise, such as
4367 1.1 mrg "A portion of the profits are donated," since it doesn't give a basis
4368 1.1 mrg for comparison.
4369 1.1 mrg
4370 1.1 mrg Even a precise fraction "of the profits from this disk" is not very
4371 1.1 mrg meaningful, since creative accounting and unrelated business decisions
4372 1.11 mrg can greatly alter what fraction of the sales price counts as profit. If
4373 1.11 mrg the price you pay is $50, ten percent of the profit is probably less
4374 1.1 mrg than a dollar; it might be a few cents, or nothing at all.
4375 1.1 mrg
4376 1.1 mrg Some redistributors do development work themselves. This is useful
4377 1.1 mrg too; but to keep everyone honest, you need to inquire how much they do,
4378 1.1 mrg and what kind. Some kinds of development make much more long-term
4379 1.1 mrg difference than others. For example, maintaining a separate version of
4380 1.1 mrg a program contributes very little; maintaining the standard version of a
4381 1.1 mrg program for the whole community contributes much. Easy new ports
4382 1.1 mrg contribute little, since someone else would surely do them; difficult
4383 1.11 mrg ports such as adding a new CPU to the GNU Compiler Collection contribute
4384 1.11 mrg more; major new features or packages contribute the most.
4385 1.1 mrg
4386 1.1 mrg By establishing the idea that supporting further development is "the
4387 1.1 mrg proper thing to do" when distributing free software for a fee, we can
4388 1.1 mrg assure a steady flow of resources into making more free software.
4389 1.1 mrg
4390 1.1 mrg Copyright (C) 1994 Free Software Foundation, Inc.
4391 1.1 mrg Verbatim copying and redistribution of this section is permitted
4392 1.1 mrg without royalty; alteration is not permitted.
4393 1.1 mrg
4394 1.1 mrg
4395 1.3 mrg File: libgomp.info, Node: Library Index, Prev: Funding, Up: Top
4396 1.1 mrg
4397 1.3 mrg Library Index
4398 1.3 mrg *************
4399 1.1 mrg
4400 1.1 mrg [index]
4401 1.1 mrg * Menu:
4402 1.1 mrg
4403 1.11 mrg * Environment Variable: OMP_CANCELLATION. (line 6)
4404 1.11 mrg * Environment Variable <1>: OMP_DISPLAY_ENV. (line 6)
4405 1.11 mrg * Environment Variable <2>: OMP_DEFAULT_DEVICE. (line 6)
4406 1.11 mrg * Environment Variable <3>: OMP_DYNAMIC. (line 6)
4407 1.11 mrg * Environment Variable <4>: OMP_MAX_ACTIVE_LEVELS. (line 6)
4408 1.11 mrg * Environment Variable <5>: OMP_MAX_TASK_PRIORITY. (line 6)
4409 1.11 mrg * Environment Variable <6>: OMP_NESTED. (line 6)
4410 1.11 mrg * Environment Variable <7>: OMP_NUM_THREADS. (line 6)
4411 1.11 mrg * Environment Variable <8>: OMP_PROC_BIND. (line 6)
4412 1.11 mrg * Environment Variable <9>: OMP_PLACES. (line 6)
4413 1.11 mrg * Environment Variable <10>: OMP_STACKSIZE. (line 6)
4414 1.11 mrg * Environment Variable <11>: OMP_SCHEDULE. (line 6)
4415 1.11 mrg * Environment Variable <12>: OMP_THREAD_LIMIT. (line 6)
4416 1.11 mrg * Environment Variable <13>: OMP_WAIT_POLICY. (line 6)
4417 1.11 mrg * Environment Variable <14>: GOMP_CPU_AFFINITY. (line 6)
4418 1.11 mrg * Environment Variable <15>: GOMP_DEBUG. (line 6)
4419 1.11 mrg * Environment Variable <16>: GOMP_STACKSIZE. (line 6)
4420 1.11 mrg * Environment Variable <17>: GOMP_SPINCOUNT. (line 6)
4421 1.11 mrg * Environment Variable <18>: GOMP_RTEMS_THREAD_POOLS.
4422 1.10 mrg (line 6)
4423 1.10 mrg * FDL, GNU Free Documentation License: GNU Free Documentation License.
4424 1.7 mrg (line 6)
4425 1.11 mrg * Implementation specific setting: OMP_NESTED. (line 6)
4426 1.11 mrg * Implementation specific setting <1>: OMP_NUM_THREADS. (line 6)
4427 1.11 mrg * Implementation specific setting <2>: OMP_SCHEDULE. (line 6)
4428 1.11 mrg * Implementation specific setting <3>: GOMP_STACKSIZE. (line 6)
4429 1.11 mrg * Implementation specific setting <4>: GOMP_SPINCOUNT. (line 6)
4430 1.11 mrg * Implementation specific setting <5>: GOMP_RTEMS_THREAD_POOLS.
4431 1.1 mrg (line 6)
4432 1.1 mrg * Introduction: Top. (line 6)
4433 1.1 mrg
4434 1.1 mrg
4435 1.1 mrg
4436 1.1 mrg Tag Table:
4437 1.11 mrg Node: Top2083
4438 1.11 mrg Node: Enabling OpenMP4489
4439 1.11 mrg Node: Runtime Library Routines5276
4440 1.11 mrg Node: omp_get_active_level8339
4441 1.11 mrg Node: omp_get_ancestor_thread_num9039
4442 1.11 mrg Node: omp_get_cancellation9969
4443 1.11 mrg Node: omp_get_default_device10783
4444 1.11 mrg Node: omp_get_dynamic11459
4445 1.11 mrg Node: omp_get_level12334
4446 1.11 mrg Node: omp_get_max_active_levels12954
4447 1.11 mrg Node: omp_get_max_task_priority13659
4448 1.11 mrg Node: omp_get_max_threads14279
4449 1.11 mrg Node: omp_get_nested15036
4450 1.11 mrg Node: omp_get_num_devices15950
4451 1.11 mrg Node: omp_get_num_procs16471
4452 1.11 mrg Node: omp_get_num_teams17010
4453 1.11 mrg Node: omp_get_num_threads17526
4454 1.11 mrg Node: omp_get_proc_bind18615
4455 1.11 mrg Node: omp_get_schedule19536
4456 1.11 mrg Node: omp_get_team_num20490
4457 1.11 mrg Node: omp_get_team_size20989
4458 1.11 mrg Node: omp_get_thread_limit21949
4459 1.11 mrg Node: omp_get_thread_num22568
4460 1.11 mrg Node: omp_in_parallel23439
4461 1.11 mrg Node: omp_in_final24088
4462 1.11 mrg Node: omp_is_initial_device24762
4463 1.11 mrg Node: omp_set_default_device25455
4464 1.11 mrg Node: omp_set_dynamic26246
4465 1.11 mrg Node: omp_set_max_active_levels27132
4466 1.11 mrg Node: omp_set_nested27909
4467 1.11 mrg Node: omp_set_num_threads28801
4468 1.11 mrg Node: omp_set_schedule29669
4469 1.11 mrg Node: omp_init_lock30750
4470 1.11 mrg Node: omp_set_lock31403
4471 1.11 mrg Node: omp_test_lock32258
4472 1.11 mrg Node: omp_unset_lock33234
4473 1.11 mrg Node: omp_destroy_lock34165
4474 1.11 mrg Node: omp_init_nest_lock34842
4475 1.11 mrg Node: omp_set_nest_lock35577
4476 1.11 mrg Node: omp_test_nest_lock36492
4477 1.11 mrg Node: omp_unset_nest_lock37519
4478 1.11 mrg Node: omp_destroy_nest_lock38534
4479 1.11 mrg Node: omp_get_wtick39285
4480 1.11 mrg Node: omp_get_wtime39877
4481 1.11 mrg Node: Environment Variables40653
4482 1.11 mrg Node: OMP_CANCELLATION42208
4483 1.11 mrg Node: OMP_DISPLAY_ENV42741
4484 1.11 mrg Node: OMP_DEFAULT_DEVICE43444
4485 1.11 mrg Node: OMP_DYNAMIC44224
4486 1.11 mrg Node: OMP_MAX_ACTIVE_LEVELS44820
4487 1.11 mrg Node: OMP_MAX_TASK_PRIORITY45470
4488 1.11 mrg Node: OMP_NESTED46128
4489 1.11 mrg Node: OMP_NUM_THREADS46733
4490 1.11 mrg Node: OMP_PROC_BIND47421
4491 1.11 mrg Node: OMP_PLACES48613
4492 1.11 mrg Node: OMP_STACKSIZE50790
4493 1.11 mrg Node: OMP_SCHEDULE51614
4494 1.11 mrg Node: OMP_THREAD_LIMIT52312
4495 1.11 mrg Node: OMP_WAIT_POLICY52912
4496 1.11 mrg Node: GOMP_CPU_AFFINITY53604
4497 1.11 mrg Node: GOMP_DEBUG55334
4498 1.11 mrg Node: GOMP_STACKSIZE55841
4499 1.11 mrg Node: GOMP_SPINCOUNT56670
4500 1.11 mrg Node: GOMP_RTEMS_THREAD_POOLS57874
4501 1.11 mrg Node: Enabling OpenACC60052
4502 1.11 mrg Node: OpenACC Runtime Library Routines61045
4503 1.11 mrg Node: acc_get_num_devices64839
4504 1.11 mrg Node: acc_set_device_type65565
4505 1.11 mrg Node: acc_get_device_type66329
4506 1.11 mrg Node: acc_set_device_num67044
4507 1.11 mrg Node: acc_get_device_num67828
4508 1.11 mrg Node: acc_async_test68625
4509 1.11 mrg Node: acc_async_test_all69617
4510 1.11 mrg Node: acc_wait70517
4511 1.11 mrg Node: acc_wait_all71379
4512 1.11 mrg Node: acc_wait_all_async72138
4513 1.11 mrg Node: acc_wait_async72890
4514 1.11 mrg Node: acc_init73597
4515 1.11 mrg Node: acc_shutdown74243
4516 1.11 mrg Node: acc_on_device74911
4517 1.11 mrg Node: acc_malloc75915
4518 1.11 mrg Node: acc_free76414
4519 1.11 mrg Node: acc_copyin76841
4520 1.11 mrg Node: acc_present_or_copyin77951
4521 1.11 mrg Node: acc_create79577
4522 1.11 mrg Node: acc_present_or_create80732
4523 1.11 mrg Node: acc_copyout82366
4524 1.11 mrg Node: acc_delete83389
4525 1.11 mrg Node: acc_update_device84364
4526 1.11 mrg Node: acc_update_self85476
4527 1.11 mrg Node: acc_map_data86580
4528 1.11 mrg Node: acc_unmap_data87265
4529 1.11 mrg Node: acc_deviceptr87786
4530 1.11 mrg Node: acc_hostptr88356
4531 1.11 mrg Node: acc_is_present88920
4532 1.11 mrg Node: acc_memcpy_to_device90447
4533 1.11 mrg Node: acc_memcpy_from_device91110
4534 1.11 mrg Node: acc_get_current_cuda_device91794
4535 1.11 mrg Node: acc_get_current_cuda_context92391
4536 1.11 mrg Node: acc_get_cuda_stream92985
4537 1.11 mrg Node: acc_set_cuda_stream93541
4538 1.11 mrg Node: OpenACC Environment Variables94073
4539 1.11 mrg Node: ACC_DEVICE_TYPE94532
4540 1.11 mrg Node: ACC_DEVICE_NUM94768
4541 1.11 mrg Node: GCC_ACC_NOTIFY95025
4542 1.11 mrg Node: CUDA Streams Usage95248
4543 1.11 mrg Ref: CUDA Streams Usage-Footnote-197148
4544 1.11 mrg Node: OpenACC Library Interoperability97257
4545 1.11 mrg Ref: OpenACC Library Interoperability-Footnote-1103613
4546 1.11 mrg Ref: OpenACC Library Interoperability-Footnote-2103865
4547 1.11 mrg Node: The libgomp ABI104073
4548 1.11 mrg Node: Implementing MASTER construct104929
4549 1.11 mrg Node: Implementing CRITICAL construct105343
4550 1.11 mrg Node: Implementing ATOMIC construct106082
4551 1.11 mrg Node: Implementing FLUSH construct106563
4552 1.11 mrg Node: Implementing BARRIER construct106834
4553 1.11 mrg Node: Implementing THREADPRIVATE construct107103
4554 1.11 mrg Node: Implementing PRIVATE clause107756
4555 1.11 mrg Node: Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses108337
4556 1.11 mrg Node: Implementing REDUCTION clause109661
4557 1.11 mrg Node: Implementing PARALLEL construct110218
4558 1.11 mrg Node: Implementing FOR construct111475
4559 1.11 mrg Node: Implementing ORDERED construct113473
4560 1.11 mrg Node: Implementing SECTIONS construct113779
4561 1.11 mrg Node: Implementing SINGLE construct114545
4562 1.11 mrg Node: Implementing OpenACC's PARALLEL construct115257
4563 1.11 mrg Node: Reporting Bugs115515
4564 1.11 mrg Node: Copying115877
4565 1.11 mrg Node: GNU Free Documentation License153424
4566 1.11 mrg Node: Funding178547
4567 1.11 mrg Node: Library Index181073
4568 1.1 mrg
4569 1.1 mrg End Tag Table
4570